INDUSTRIAL PERFORMANCE CENTERMassachusetts Institute of Technology

Cambridge, MA 02139

INSTITUTIONS AND INDUSTRIAL PERFORMANCE:THE PHARMACEUTICAL SECTOR IN FRANCE,

GERMANY, BRITAIN, AND THE U.S.1

J. Gunnar Trumbull

MIT IPC Working Paper 00-0002MIT IPC © 2000.

The views expressed herein are the authors’ responsibility and do notnecessary reflect those of the MIT Industrial Performance Center or theMassachusetts Institute of Technology.

1 This research has been supported by the Industrial Performance Center at the MassachusettsInstitute of Technology.

2

INSTITUTIONS AND INDUSTRIAL PERFORMANCE:THE PHARMACEUTICAL SECTOR IN FRANCE,

GERMANY, BRITAIN, AND THE UNITED STATES2

Table of Contents

I. Varieties of Capitalism and National Product Market Strategies.................... 3II. Indicators of Pharmaceutical Industry Performance ....................................... 5

Industry’s Perspective on Pharmaceutical Performance.................................... 6Governments’ Perspective on Pharmaceutical Performance........................... 10Summary of Performance Indicators................................................................. 18

III. Government Regulation of Drug Quality and Price....................................... 23Germany: Regulatory Laxity .............................................................................. 26Britain: Regulation via Negotiation .................................................................... 29France: Price Regulation ................................................................................... 31United States: Quality Regulation ..................................................................... 36

IV. Conclusions.................................................................................................... 39Reference List .................................................................................................... 42

2 This research has been supported by the Industrial Performance Center at the MassachusettsInstitute of Technology.

3

I. Varieties of Capitalism and National Product Market Strategies

The Varieties of Capitalism (VOC) approach to political economy assertsthat domestic producers are able to secure a competitive advantage in the globaleconomy by pursuing production strategies that emphasize the institutionaladvantages that for historical reasons are available within their countries. ThisVOC analysis arose as a reaction both against the prior focus on aggregatedemand conditions emphasized by Keynesian analysis (Streeck 1992, 2), as wellas in response to trade theories targeting aggregate factor endowments as asource of comparative advantage (Hollingsworth 1997, 283). Advocates of VOCargue that nationally distinctive institutional contexts of production have self-reinforcing mechanism that make them likely to persist over time. Theinstitutional configuration of labor training and supplier contracting “function asconstraints and opportunities simultaneously,” at once raising the cost ofalternative product market strategies and creating a comparative advantage inexisting strategies (Streeck 1992, 29). This comparative advantage argumentimplies that globalization in trade creates competitive forces that perpetuate andintensify national institutional distinctiveness. Distinctive institutional forms shouldtherefore play a central role in setting patterns of trade among the advancedindustrial countries. Thus the VOC model offers an account both of the sourcesof prevailing national product strategies, and of the patterns of globalization thatare likely to emerge among countries with similar factor endowments (land, labor,capital) but organized through different national institutions. I test the predictionsof the VOC approach against the experience over the past thirty years of thepharmaceutical industry in Germany, France, Britain, and the United States.

Researchers working in the VOC mode have generated a range ofsubstantially consistent theories linking the institutional form of production toproduct market strategies. David Soskice has focused on nationally specificconfigurations of the financial system, the industrial relations system, theeducation and training system, and the inter-company system. He finds thatvariation in the institutional configuration of these four systems falls into twobroad production strategies: liberal market economies and coordinated marketeconomies (Soskice 1996a, 102-5). Coordinated market economies pushmanufacturers toward a product market strategy that emphasizes high-qualityproducts and incremental innovation. Liberal market economies foster productmarket strategies that emphasize lower quality and more radical forms ofinnovation (Soskice 1996b). Working in a similar framework, Wolfgang Streeckfocuses on worker skill formation and the nature of supplier relations. Whereworker skills are high and supplier relations congenial, as in Germany,companies excel in product market strategies that emphasize diversified qualityproduction (Streeck 1992, 4). Streeck argues, for example, that the institutionalcontext of diversified quality production in Germany has kept German automobilemanufacturers from importing Japanese-inspired strategies of lean production(Streeck 1996).

4

This paper tests the hypotheses emerging from the VOC model againstnational performance in the pharmaceutical sector in France, Germany, Britain,and the United States. In both the Soskice and the Streeck models, Germanyoccupies one product-market extreme while the United States and Britain bothoccupy another. Because the pharmaceutical industry is innovation-intensive, theVarieties of Capitalism approach suggests that the United States and Britainshould excel in this sector, especially in relation to new kinds of products such asthose developed through bio-engineering. A recent comparative study of theBritish and German pharmaceutical sectors by Steve Casper and CatherineMatraves (1997) supports this hypothesis.

While the United States is closely associated with Britain under the liberalmarket economy category, the place of France in this schema remains disputed.Michel Albert (1991) argued in the early 1990s that France teeters between theseductive allure of the Anglo-American model of liberal market economy and themore prudent austerity of the Rheinland model of coordinated market economy.Jonah Levy (1999) cautions that France simply lacks the institutionalinfrastructure — in the form of social capital, associational density, anddecentralized control mechanisms — to pursue the Rheinland modelsuccessfully. Alternatively, Bob Hancké (1999) has proposed that, even in theabsence of such associational capacities, France has nonetheless been able toadopt coordinated industrial strategies by taking advantage of state-centeredelites capable of promoting high-end industrial strategies through government-managed labor and capital policies. Nick Ziegler’s research on French elites findsthat the French approach may only succeed in certain kinds of network industriesbut not in industries that require decentralized reforms (Ziegler 1997). Iprovisionally include France in the category of coordinated market economies,with the underlying assumption that France will not enjoy a comparativeadvantage in the innovation-intensive pharmaceutical market.

The paper presents its findings in two sections. The first section tests VOChypotheses against industry performance indicators for France, Germany,Britain, and the United States. Performance is measured first from theperspective of industry itself, focusing on profitability, sales, and foreign marketpenetration. A second measure of performance focuses on indicators ofparticular interest to national governments: employment, balance of trade, totaldrug bill, and innovation. Where possible, I present historical performance data inorder to place current trends in a longer perspective. These data show that theVOC model explains some but not all of the variation in national competitivenessin the pharmaceutical industry. As predicted, Britain does out-perform Germanyin pharmaceutical innovation and export. Yet Germany also out-performs theUnited States in a number of indicators, a finding not expected under the VOCmodel.

The second section of the paper traces national approaches to regulatingdomestic pharmaceutical markets, focusing first on safety and efficacy regulation,then on price regulation. How countries have regulated pharmaceutical productsalong these dimensions appears to have played a role in setting industry productstrategies. In France, where direct price restraint has been unusually

5

comprehensive, pharmaceutical firms have sacrificed new drug development infavor of a low-cost, high-volume strategy of domestic sales. In Britain, industryprofits are negotiated between producers and the government based on theresearch and export intensity of individual firms. British firms have thereforeemphasized R&D and export activities, while de-emphasizing their domesticsales effort.

This paper confirms many of the predictions of the Varieties of Capitalismapproach. But it also finds that a more thorough explanation of product marketstrategy in the pharmaceutical industry must incorporate the institutions that setnational conditions of product demand. Recent research in businessmanagement has suggested that demand conditions may be importantdeterminants of industry strategy (Porter 1990, 87; Kogut 1991, 36; Albach 1994,269; Storper and Salais 1997; Gholz 1999, Chapter 4). One cross-national studyof the impact of different kinds of regulation, for example, finds that productmarket regulations are nearly twice as important to national economic prosperityas are labor market regulations (Koekijk et al. 1996).

This emerging body of research suggests that national constellations ofproduct market regulations may have a strong impact on industry strategy. Inmany cases these product market effects may complement the impact ofdistinctive national production systems as described in the Varieties of Capitalismliterature. But the forms of national product market regulation do not necessarilymirror national production institutions in a simple way. In the Britishpharmaceutical industry, for example, the government assigns profit targets formanufacturers based on their research investment and export success, and thenimposes drug prices such that each firm’s profitability target is met. The approachlooks decidedly illiberal with respect to the economic autonomy of the firm, andtherefore antithetical to the liberal market economy ideal type. But it neverthelessserves to bridge the potentially conflicting interests of shareholder value andgovernment budget restraint. This example suggests how a broader approach tothe Varieties of Capitalism analysis, one that incorporates national configurationsboth of production institutions and of consumption institutions, may help toexplain important dimensions of observed variation among national industrystrategies.

II. Indicators of Pharmaceutical Industry Performance

This section tests the VOC hypotheses against aggregate indicators ofindustrial performance for the pharmaceutical industries of France, Germany,Britain, and the United States. Because industrial performance is often in the eyeof the beholder, we consider first industry-oriented standards of performance,then government-oriented standards of performance. We then evaluate theimpact of supply-side institutions, those conventionally addressed in the VOCliterature, on differential performance in the four test countries. We conclude thatthe VOC approach, as currently deployed, helps to explain some but not allnational variation in industrial performance.

6

The VOC approach makes two broad predictions about how thepharmaceutical industry will perform in different countries. In those dominated bycoordinated markets, such as Germany and France, we should expect to seepoor industry performance due to an inability to sponsor the kind of radicalinnovation required for success in the pharmaceutical industry. By contrast, weshould expect a high degree of success by British and US pharmaceutical firms.The second and related prediction of the VOC model is that these differentialnational capabilities should drive patterns of pharmaceutical globalization. Britainand the US should enjoy a strongly positive balance of trade in pharmaceuticals.France and Germany should experience a correspondingly weak balance oftrade. We should also see signs that pharmaceutical producers in thecoordinated market economies are moving research and production activities toliberal market economies that offer greater advantages for innovation.

How do the pharmaceutical industries in these countries measure up? Theevaluation proceeds in two parts. The first section below takes the perspective ofindustry, and asks how each national industry is performing by comparison withproducers in other nations. Here we look at the bottom line: profits, volume ofproduction, exports, and ownership. The second section takes the perspective ofthe national government. It asks what costs and benefits the pharmaceuticalindustry is bringing to consumers, to the government, and to the larger economy.This section looks at the national drug bill, the trade balance of the drug industry,the employment opportunities it provides, and the extent to which pharmaceuticalfirms invest in health-improving research. Where possible I provide historicaldata, which permits us to see how current patterns have emerged over time. Formost indicators I also present a per capita measure for the latest year available.While these per capita measurements often convey little useful meaning on theirown, they do permit us to account for the very different sizes of the casecountries, especially the disparity between the United States and the smallerEuropean countries.

Industry’s Perspective on Pharmaceutical Performance

This section focuses on performance indicators of interest to managersand owners in the pharmaceutical industry. For owners, the main interest isreturn on investment, and therefore company profits. For managers, success isdefined by profits, but also by sales volume, export success, and in part also byacquisition of foreign firms.

Profitability data on pharmaceutical manufacturers strongly affirm the VOChypothesis about the relative strength of liberal market economies in an industrycharacterized by radical innovation. British pharmaceutical firms enjoy aparticularly high level of profit, which, at 28.7 percent, is nearly twice as high asprofitability in the United States. France, at 5.8 percent, has the lowestprofitability of the four. In Germany, profits, at 11.3 percent, are greater than inFrance but less than in the United States or Britain. This pattern of national profitmargins is consistent with the central role attributed to the stock market in liberalmarket economies, with the dominant role of longer-term sources of investment

7

finance in coordinated market economies, and with the hypothesized innovativeadvantage of liberal market economies.

Figure 1. Pharmaceutical profit margin, percent of revenue.

1983 1991

France 7.8 5.8

Germany 13.8 11.3

Britain 32.1 28.7

US n/a 15.2Source: Burstall and Reuben 1987, 120; Finkelstein and Bittinger 1993, 4-5.

Production figures, however, complicate the picture. Frenchpharmaceutical companies, taken together, enjoy higher sales than do Germanor British pharmaceutical firms. While US firms outsell all others by a wide margindue to the large size of the country, a comparison of sales measured on a percapita basis reveals that, relative to the size of the economy, Frenchpharmaceutical producers also far outsell manufacturers in the United States.From this perspective, German sales are the lowest of all four countries. It isinteresting to note that as recently as 1990, Germany produced morepharmaceuticals than did France or Britain. Indeed in 1983, Germany producedtwice as many drugs as Britain and half again as much as France. This sign ofstrong sales by France and, earlier, Germany, is surprising given ourexpectations about the relative capabilities of liberal and coordinated marketeconomies.

Figure 2. Pharmaceutical production, $ billions.

1961 1970 1983 1990 1997 1997 in $per capita

France 0.6 1.4 7.9 14.1 22.3 381

Germany 0.6 1.7 11.2 15.9 19.8 241

Britain 0.5 0.7 5.6 11.9 17.5 297

US 3.2 6.8 23.9 41.9 75.8 284Source: ABPI; BPI; Matraves 1998, 26; Möbius et al. 1976, 101; PhRMA; SNIP.

The record of export success of pharmaceutical producers in thesecountries holds similar surprises. Most notably, the volume of US drug exports islower than those of France, Germany, or Britain, despite the substantially largersize of the US economy. Indeed, the historical trend shows that the United Stateshas lagged behind the European countries in export volume since the 1970s. In

8

terms of actual volume, the most successful drug exporter has long beenGermany, which exports 50 percent more pharmaceutical products than does theUnited States. For France, strong production figures have not translated into astrong export success (although France does exceed the United States in exportvolume). This suggests, as we will see below, that the French themselvesconsume a larger volume of drugs, measured on a per capita basis, than do theother countries studied. Taking the 1997 figures on a per capita basis, Britainemerges as the best export performer, followed closely by Germany and thenFrance. This measure indicates just how poorly the United States performs indrug export compared to the other countries. One of the reasons for this is, as wewill also see below, that US pharmaceutical companies have favored producingoverseas rather than exporting finished products.

Figure 3. Pharmaceutical exports, $ billions.

1961 1970 1980 1990 1997 1997 in $per capita

France 0.10 0.16 1.09 2.85 8.5 145.04

Germany 0.14 0.32 1.28 3.61 12.7 154.78

Britain 0.15 0.25 1.13 3.14 9.6 162.69

US 0.27 0.18 0.66 1.33 7.9 29.61Source: Industrie pharmaceutique 1996, 36; ABPI.

Taking exports as a percentage of total sales (see Figure 4 below), wesee that Germany and Britain lead France and especially the United States inexport intensity. Once again, we observe a pattern of industrial performance that,by cutting across the coordinated/liberal market division, confounds the VOChypothesis about success in drug trade. Note the extraordinary growth inGerman exporting over the course of the 1990s, during a time when theintroduction of radically new technologies, especially biotechnology, wereexpected to put Germany at a disadvantage. Notice also that the overall exportintensity of pharmaceutical producers has grown importantly in all four countries(doubled in France and Britain, tripled in Germany and the United States) overthe past decade. Finally, note the drop in trade intensity from 1961 to 1970,which reflects the impact of a period of intensive safety regulation for thepharmaceutical industry that followed in the wake of the Thalidomide (Contergan)tragedies. As these trade figures show, the export intensity of the pharmaceuticalindustry has increased dramatically only beginning in the 1990s.

9

Figure 4. Export as a percentage of total sales.

1961 1970 1990 1997

France 16.7% 11.4% 20.2% 38.1%

Germany 23.3% 18.8% 22.7% 64.1%

Britain 30.0% 35.7% 26.4% 54.9%

US 8.4% 2.6% 3.2% 10.4%

One reason for the relatively low export intensity of US producers is theirstrategy of developing and manufacturing pharmaceuticals in their targetmarkets. US drug companies have bought or established a large number ofproduction sites in France, Germany, and Britain. Likewise, producers in thesecountries also own drug companies in the United States and in other Europeancountries. The trade data presented above are based on the geographicallocation of a company, not on its ownership. But ownership matters to companymanagement, and especially to shareholders, and thus must be considered inour assessment of industrial performance. Figure 5 below describes the share ofdrug sales in each country according to the nationality of the producers. Shadedsquares indicate the percentage of national consumption served by nationalpharmaceutical producers. As these data indicate, only in the United States dodomestically-owned companies provide more than half of the drugs consumed inthe country. In Britain, incredibly, only 26 percent of domestic consumption isprovided by British pharmaceutical firms. France and Germany are in anintermediate position, with 33.8 percent of the French market served by Frenchfirms, and 46.6 percent of the German market served by German firms. Thesefigures attest to an astounding degree of globalization in ownership in thepharmaceutical industry.

Figure 5. Share of country market consumption according to the nationalityof producers, measured as percentage of domestic consumption, in 1997.

country markets

Germany France Britain USA

German 46.6 12.9 8 7.5

French 5.1 38.8 4.6 1.5

British 6.1 8.2 25.8 12.3

US 19.7 23.6 31.3 63.9na

tion

alit

y o

fp

rod

uce

rs

all other 22.5 16.5 30.3 14.8Source: SNIP.

10

US firms have extended heavily into German, French, and especiallyBritish markets. Nearly a third of all drugs sold in Britain come from USproducers. Britain, in turn, has been the most successful of the Europeancountries in penetrating the United States market, with 12.3 percent of marketshare. Germany has captured 7.5 percent of the US market. France has onlymanaged to take over 1.5 percent of the US market. Success in the US market isparticularly significant for European firms because the market is so large. Thus itprovides a useful indicator of the relative market penetration of European firms(Casper and Matraves 1997, 2). From this we conclude that the USpharmaceutical industry has been most successful at penetrating foreignmarkets, followed by Britain, Germany, and, in last place, France.

The picture that emerges from these industry-oriented performanceindicators is inconsistent with respect to the VOC model. Profitability figuresreinforce VOC claims about the different innovative strengths of liberal andcoordinated market economies. Figures on the saturation of foreign marketsreinforce the view that British and US firms are most successful. But productionfigures, especially when measured on a per capita basis, show that France sellsmore pharmaceuticals than do the United States or Britain. This finding issurprising in a coordinated market economy. Equally surprising is the distributionof export success of the four countries studied. In a pairing that cuts directlyacross the VOC categorization, British and German pharmaceutical industriesare export intensive.

Governments’ Perspective on Pharmaceutical Performance

Before addressing these apparent inconsistencies we must also considerthe performance of the pharmaceutical industry as viewed from the perspectiveof national governments. National interests include employment, trade balance,cost to society, and innovative capacity, which in the long term promises togenerate more effective medicines.

Measured in terms of employment, Germany’s pharmaceutical industry isa strong performer, with 114,000 pharmaceutical employees in 1997. See Figure6 below. France lags behind Germany, with 73,000 employees. Counted perthousand of population, France falls only slightly behind Germany in the overallcontribution of pharmaceuticals to employment. Britain, by contrast, has thesmallest number of pharmaceutical workers, and indeed had to reduce the workforce by 11,000 between 1990 and 1997. Britain lags behind Germany andFrance both in absolute employment figures and as a portion of population. TheUnited States, which has the largest number of pharmaceutical employees,nonetheless has the smallest workforce in relation to the country’s population.Hence the European countries in general, and France and Germany especially,have been successful in sponsoring employment in the pharmaceutical sector.For France, this high level of employment is consistent with the very high sales inthe country. For Germany, however, where per capita production is the lowest ofall four countries, the high employment suggests a low labor productivity. Hence,the performance of the French and German pharmaceutical industries on theemployment scale is more favorable than the performance of the British or US

11

industries. This finding is consistent with the profitability figures above, and withthe expectations of the VOC model, since excessive employment, enforcedthrough rigid labor market rules, works to restrain productivity growth.

Figure 6. Pharmaceutical sector employment, thousands.

1975 1983 1990 1997 1997 perthousand

France 56 65 66 73 1.25

Germany n/a 87 100 114 1.39

Britain 67 67 71 60 1.02

US n/a 153 189 208 0.78Source: ABPI; BPI; European Community 1997, 65-66; PhRMA; SNIP

National governments also have an interest in the trade balance ofindustry, preferring in general to export more than they import. While tradedeficits are not in themselves bad, an overall negative balance of trade forces acountry to finance excess imports through borrowing, which in turn increases thecost of government borrowing. Trade balance figures for 1997 show thatGermany exports far more pharmaceuticals than it imports. This is also true, to alesser extent, for Britain and for France. The United States in 1997 ran a tradedeficit in pharmaceuticals. Given that liberal market economies are expected toexcel in the pharmaceutical sector, poor US trade performance runs contrary tothe VOC model. The poor US trade performance is due largely to the decision ofUS producers to develop and manufacture drugs for foreign markets in the targetcountries.

Figure 7. Trade balance in 1997, $ billions.

exports imports balance

France 8.5 5.7 2.8

Germany 12.7 6.8 5.8

Britain 9.6 5.6 4.0

US 7.9 9.3 -1.4Source: ABPI.

The pharmaceutical industry is unusual in that governments pay, directlyor indirectly, much of the consumer’s bill. This gives governments a stronginterest in lowering per capita drug consumption, whether the drugs come fromdomestic producers or from foreign sources. The interest of the government inlower overall drug consumption should therefore depend on how much of the

12

total national drug bill they are expected to pay. The government burden issignificantly different in the four countries we study. At the extremes, the BritishNational Health Service pays for nearly 90 percent of the national drug bill, whilethe United States government pays for only about 40 percent. The French andGerman governments each pay just over 60 percent of the total cost of drugs(Elgar et al. 1992, 47). This suggests that the government of the Unites Stateshas the least budgetary incentive to keep drug spending low. This costindifference is indeed reflected in a relatively high level of spending on drugs,averaging 319 dollars per person in the United States in 1997. By the same logic,the high burden of payment of the British government is reflected in an unusuallylow per capita drug cost, only 233 dollars per person in the same year.Consistent with this analysis, Germany’s drug bill has an intermediate valuebetween the United States and Britain. The surprise is France, which despite arelatively high government contribution to drug payment, has experienced thelargest drug bill of all four countries. This finding is consistent with, and indeedmust be considered the proximate cause of, France’s large domesticpharmaceutical production, noted above. The French are, in fact, the biggestdrug users in the world, in per capita consumption, followed in close second byJapan (PhRMA 1999). This high level of domestic consumption may be good forFrench producers but it is bad for the French state, which pays much of the bill.

It is interesting to note the historical trend in national drug consumption.First, Britain has always, or at least since the early 1960s, consumed fewer drugsthan its counterparts. France, by contrast, has not always been the biggestconsumer. Prior to the mid-1970s, consumers in the United States spent more ondrugs than did French consumers. In fact the surge in French drug consumptionemerged only in the 1990s. French drug spending grew 63 percent from 1991 to1997, during a period of relatively low inflation. Britain experienced an evengreater surge during this period, of 120 percent. This rapid growth in part reflectsthe very low level of drug consumption in 1991, when British consumers usedless than half as many drugs, measured by cost, as consumers in the remainingthree countries. But the rapid growth trend has led British officials to worry todayabout drug spending despite the current low cost in Britain relative to othercountries. By comparison, German spending grew by only 34 percent over thistime, and US spending by 49 percent. As we will see below, these spendingtrends, which have been formative for national industries, have their source notmerely in cultural orientations toward medicine, but also in distinctive strategiesof product market regulation adopted by the national governments.

13

Figure 8. Per capita drug consumption, $.

1961 1970 1979 1991 1997

France 11.4 25.3 81.1 215 351

Germany 8.4 23.2 81.5 220 294

Britain 6.8 8 35.8 101 233

USA 16.4 31.6 51.5 214 319Source: Burstall 1985, 10; Faibus 1984, 5; Finkelstein et al. 1993, 25; Möbius1976, 103; PhRMA.

Finally, governments have an interest in the long-term health of theircitizens and of their drug industries, and therefore place a high value on researchinto new kinds of drugs that may benefit society in the future. Because drugresearch is a chancy business, with the fortunes of companies hanging on thesuccess or failure of single compounds, we consider both effort and success inpharmaceutical research. Effort is measured in terms of spending andemployment in research and development. Success is measured in terms of thenumber of new discoveries, specifically new drug substances. Both indicatorsconfirm VOC expectations about the innovation-emphasis of liberal marketeconomies.

Considering research and development (R&D) effort, Britain and theUnited States systematically outspend France and Germany. See Figure 9below. In both absolute and per capita terms, British and US pharmaceuticalfirms spend more on R&D than French and German firms. It is interesting tonote, however, that France and the United States spend nearly the sameamount, per person, on R&D. Germany spends far less per person on R&D thando the other three countries. This findings is surprising given the strong exportorientation of the German industry, but it is consistent with the VOC view thatradical innovation is not an area of comparative advantage for coordinatedmarket economies.

Figure 9. Pharmaceutical R&D spending, $ millions.

1964 1973 1980 1991 1996 1996 $per capita

France 40 173 261 1,880 2,794 48*

Germany 27.5 299 990 2,330 3,136 38

Britain 29 145 292 2,120 3,206 55

USA 282 828 1,242 6,750 13,230 50Sources: BPI; Finkelstein 1993, 4-5; Industrie pharmaceutique 1995, 9; *1995Möbius 1976, 120; OECD 1985, 30; SNIP.

14

Another way to evaluate R&D intensity is as a percentage of total sales.See Figure 10 below. These figures confirm the research-orientation of Britishand US pharmaceutical firms. In 1996, British and US firms invested almost 50percent more of total sales revenue into R&D activities than did French andGerman firms. The groupings are surprisingly strong. Moreover, this pattern ofR&D investment has been consistent since at least the early 1980s. Note theoverall growth in the research intensity of the pharmaceutical industry over thepast 35 years. In France and the United States, the portion of sales revenue re-invested in R&D has grown almost three times since 1964; in Germany and inBritain, R&D intensity has grown over four times in the same period.

Figure 10. Pharmaceutical R&D spending as a percentage of sales.

1964 1973 1983 1993 1996

France 4.8 7.9 7.1 12.5 14*

Germany 3.2 9.2 8.4 14.2 14.1

Britain 4.8 13.4 11.7 17.1 20

USA 7.2 9.4 10.6 16.7 19.9Sources: Industrie pharmaceutique 1995, 9; Möbius 1976, 101, 120; *1995OECD 1985, 30; SNIP.

Finally, patterns of employment in research and development reinforce thefindings from R&D spending. In France and Germany, 14 and 13 percent ofemployees work in R&D, respectively. In Britain, with the strongest R&D effort,fully one third of the workforce is focused on R&D, over twice as much as inFrance or Germany. In the United States, one quarter of employees work inR&D. Comparing these employment figures with R&D spending, it is interestingto note that R&D activities are more capital intensive in France and Germanythan in Britain or the United States. In France and Germany, R&D accounts forroughly 14 percent of employees and 14 percent of revenues. In Britain, R&Daccounts for 33 percent of employees but only 20 percent of revenues. Thetendency is repeated but less marked in the United States, where R&D accountsfor 25 percent of employees and 20 percent of revenue. This finding appears toreconfirm the VOC insight that radical innovation in coordinated marketeconomies is more difficult and therefore, measured per employee, moreexpensive. It appears to cost more to run an R&D lab in France or Germany thanit does in Britain or even the United States.

15

Figure 11. R&D employees as a percentage of total employment, 1975-1997.

1975 1983 1990 1997

France 11% 12% 17% 14%

Germany n/a 13% 15% 13%

Britain 15% 17% 26% 33%*

US n/a 26% 23% 25%Sources: ABPI; BPI; European Community 1997, 65-66; PhRMA; SNIP. *1998

Why do British and US pharmaceutical firms pursue more R&D than dotheir French and Germany counterparts? Is it because, as the VOC modelsuggests, they face a comparative disadvantage in this area of radicalinnovation? One indication is that R&D spending is very closely linked tocompany profitability. Figure 12 below shows R&D investment, measured as apercentage of revenue, plotted against profit, also measured as a percentage ofrevenue. Data for 1983 and 1991 are included. For each year, we observe astriking correlation between profit and R&D investment. Comparing the two yearsshows how dramatically the emphasis in R&D has increased in thepharmaceutical industry as a whole in the intervening 8 years. Hence we see abroad trend towards greater R&D investment over time, plus a consistent trendlinking greater R&D investment to greater profit.

16

How are we to interpret this finding? Does greater R&D investmentgenerate greater profits? Or do greater profits make more R&D investmentpossible? If the former is true, this provides strong support for the VOC model.Germany and France, in this interpretation, enjoy a comparative disadvantage inpharmaceutical innovation. They therefore invest less in R&D, with the long-termconsequence of lower profits. While this scenario seems plausible, I show in thefollowing section that it is mostly incorrect. For at least two of the countriesconsidered, France and Britain, profit levels for industry are strongly influencedby government regulation.

Given the differences in national R&D efforts, how successful have thesecountries been in finding new drugs? Innovation indicators are notoriouslyunreliable for cross-country comparisons. This is particularly true of the mostcommon indicator of new drug development, the number of new drugsubstances, or new chemical entities (NCEs), that are introduced each year.Moreover, uncertainty in the discovery process generates strong nationalvariation, even with new drug discoveries grouped into five-year periods. Butgiven these caveats, broad trends in NCE data can be instructive. See Figure 13

Figure 12. Pharmaceutical profitability and R&D spending as a percentage oftotal sales, 1983 and 1991.

0

5

10

15

20

25

30

0 10 20 30 40

Profitability (%)

R&D investment (%)

UK‘91

US‘91

France‘91

Germany‘91

France‘83

Germany‘83

UK‘83

Source: Burstall and Reuben 1987, 120; Finkelstein and Bittinger 1993, 4-5;Matraves 1998, 26.

17

below. First, and most clearly, the US pharmaceutical industry has generatedmore new drug substances than the other three countries in each of the five yearperiods. This follows from the large size, and consequently large R&Dinvestment, in the United States. Second, Britain has only contributed a highnumber of new drug compounds in the last period, from 1990 to 1994. Thisfinding is surprising given the important resources that British firms have investedin R&D activities. Finally, France and Germany have fluctuated strongly in theirsuccess in finding new drug substances. French performance today, however,falls far behind a prolific period of new drug discovery prior to 1974, a time whenFrance came very close to the level of discovery of the much larger UnitedStates.

Figure 13. New drug substances by country of discovery, 1965-1994.3

1965-69 1970-74 1975-79 1980-84 1985-89 1990-94

France 92 78 37 8 49 14

Germany 48 39 50 32 4 22

Britain 20 14 17 6 12 24

USA 93 87 71 46 90 85Source: SNIP data.

How do these unlicensed innovations translate into market success?Figure 14 below presents the cumulative record of new drug substances duringthe twenty-year period from 1975 to 1994, along with various measures ofsuccess in the international marketplace. The fundamental lesson from thesedata is that new drug substances do not translate in an obvious way into drugsthat are competitive on the world market. In the clearest case, France andGermany have generated the same number of new drug substances over thisperiod, yet twice as many of German drugs became successful in all seven majornational markets. Moreover, Britain discovered very few new drug substances, asnoted above, but those discovered were highly successful in the internationalmarket. Britain exports nearly twice as many drugs as Germany, and four timesas many drugs as France, to the seven major world markets.

3 Also called new chemical entities (NCE) or new molecular entities, these are new therapeuticdrugs in dosage form that have not yet been submitted for regulatory approval. They may consistof new complexes, simple esters, or salts of already approved drugs (FDA definition).

18

Figure 14. Cumulative innovation indicators, 1975-1994.

new drugsubstances

drugs selling in4 of the 7 major

markets4

drugs selling inall 7 majormarkets

Innovative andglobal (7 major

markets)

Innovative andGlobal products

(weighted formarket size)5

France 108 24 5 3 12

Germany 108 38 11 8 30

Britain 59 14 21 11 96

US 292 90 68 30 30Source: Finkelstein and Bittinger 1993, 3; Industrie Pharmaceutique 1996, 28.

Data for the United States shows signs both of success and of failure. Onthe one hand, the United States produces more successful new drugs than anyof the other three countries. Indeed it produces more than the other threecountries combined, by any measure. On the other hand, because the UnitedStates is so much larger than individual European states, it dedicates moreresources to innovation and should therefore be expected to generate moreinnovative products. Finkelstein and Bittinger (1993, 3) attempt to weight acombined innovation and global sales indicator in order to offset the differentialimpact of domestic market size. This index, presented in the last column ofFigure 14, finds that the United States and German perform equally well in termsof marketing new innovations. Britain, on this scale, performs extremely well,earning triple the score of Germany and the United States, despite its lowperformance in discovering new drug substances. France performs particularlypoorly on this weighted scale, earning only 12 points.

Summary of Performance Indicators

What conclusions can we draw, based on these diverse indicators, aboutthe relative industrial performance of the pharmaceutical sectors in thesecountries? One way to summarize the individual findings, including thoseprimarily of interest to industry and those primarily of interest to the government,is by employing an ordinal ranking of performance on each indicator. In Figure 15below, the countries are ranked from 1 to 4 based on their most recentperformance on each indicator, discounting, where possible, for the effect ofcountry size. A score of 4 indicates highest rank, a score of 1 lowest rank. Samelevels of performance are scored the same. The last column indicates whethercountry rankings are consistent with VOC predictions about patterns ofpharmaceutical innovation. A plus sign ( + ) indicates support for the VOC model;

4 Britain, Canada, France, Germany, Italy, Japan, and United States.5 In this scale, proposed by Finkelstein and Bittinger (1993, 3), the number of global andinnovative products is weighted to offset the effect of domestic market size. The weighting factorsare: France, 4.1; Germany, 3.7; UK, 8.7; and US, 1.

19

a question mark ( ? ) indicates ambiguous findings; a minus sign ( – ) indicatescontradiction of the VOC model.

Figure 15. Summary of Pharmaceutical Industry Performance in France,Germany, Britain and the United States.

France Germany Britain US VOC

Industry interest

profitability 1 2 4 3 +

production 4 1 3 2 ?

export6 2 3 4 1 ?

foreign penetration 1 2 3 4 –

score 8 8 14 10

Government interest

employment 3 4 2 1 +

trade balance 2 4 3 1 ?

drug bill 1 3 4 2 n/a7

R&D spending8 2 1 4 3 +

successful innovation9 1 2.5 4 2.5 +

score 9 14.5 17 9.5

Several of the indicators of industrial performance support the VOC modelas applied to the pharmaceutical industry. Drug companies in the liberal marketeconomies (Britain and US) are more profitable than those in the coordinatedmarket economies (Germany and France). As predicted, the US and Britainappear to invest more in research and development than Germany and France,and that investment appears, on the whole, to generate more useful innovations.Finally, the higher level of employment in the pharmaceutical industries in Franceand Germany is consistent with more rigid labor market regulations that

6 This refers to per-capita drug exports, not export intensity, since the measure of export intensityrewards lower levels of domestic sales with a higher score. This does not correctly reflect theemphasis of an industry-oriented performance evaluation.7 The VOC model makes no predictions about the relative level of consumption in the differentcountries.8 This refers to per capita R&D spending. This I take to be the fundamental government interest.In any case, this ranking is quite similar to rankings for R&D employment and for R&D investmentas a percentage of revenue.9 This refers to the Finkelstein/Bittinger weighted scale of innovative and global drugs. Becausewhat matters from the perspective of the government are only drugs that are effective, I do notinclude a ranking of new drug substances.

20

discourage labor mobility and may in turn hurt innovation and profitability. Thehigher employment and lower profitability of pharmaceutical industries incoordinated market economies follows directly from the configuration of labor andcapital markets in these countries. More rigid labor market discourage work-forcereduction. The predominance of bank lending and institutional ownership reducespressures for profitability.

But there are also several indicators for which the VOC model does notappear to explain observed patterns of national performance. These relateprimarily to the kind and market experience of actual pharmaceutical products. Inproduction, trade, and trade balance indicators, the liberal market economiesshow uneven performance. Whereas Britain performs well, the United Statesperforms poorly. The coordinated market economies, by contrast, performsurprisingly well. France enjoys the largest per capita production of the four.Germany enjoys the strongest trade balance of the four. For one indicator,foreign penetration, we find outcomes that appear to contradict expectationsbased on the VOC model. British and US firms have moved a lot of theirproduction to other countries. British firms appear to have concentrated in theUS. But US firms have also moved production to a surprising extent intocontinental Europe. Indeed the negative trade balance of the United States inpharmaceuticals is due to an important degree to this trend towards overseasproduction. From the VOC perspective, this trend appears unusual, since theUnited States should offer comparative advantages in drug development. I willargue below that these apparent contradictions of the VOC approach can readilybe explained by incorporating demand institutions, especially product marketregulation, into the VOC analysis.

We can also aggregate the performance indicators in order to comparebroad national experiences of pharmaceutical production. By tallying the rankingsin Figure 15 above, we can construct an admittedly blunt set of indicators ofindustry and government perceptions of success. Plotting cumulative rankings foreach country, we depict graphically the success of governments and of industryin achieving their productivity objectives. See Figure 16 below. It should be notedthat the plotted points represent only relative and not absolute values, and thatthey are highly skewed, since we are considering some of the most successfulcountries in terms of pharmaceutical production. With these caveats, however,the findings are interesting in two respects.

21

Figure 16. Plot of government and industry indicators of success.

First, they offer an criticism of the French and US pharmaceuticalindustries. Not only has the French industry failed to excel by reasonableindustrial standards of performance, it has also failed to excel by reasonablegovernment standards of performance. Perhaps more surprising is theperformance of the US pharmaceutical industry, which surpasses the French byonly a small margin. Measured according to government criteria, USperformance is nearly as poor as French performance. The Britishpharmaceutical industry is the overwhelming success story. Not only does itscore well from industry’s perspective, it also scores extremely well from thegovernment’s perspective. Germany, finally, is a mitigated success. Seen fromthe government’s perspective, pharmaceutical performance is good, if not asgood as in Britain. But from industry’s perspective, performance is poor.

Second, we see that pharmaceutical producers in liberal marketeconomies (Britain and United States) do perform better than producers incoordinated market economies (Germany and France) when we look atperformance from the perspective of industry. This is an important confirmation ofthe VOC hypothesis. Seen from the government’s perspective, however, theresults are again ambiguous. Germany and Britain sit on opposite sides of allVOC classifications, yet both do far better than the United States and France insatisfying government goals for the pharmaceutical industry. This finding is notsurprising in the case of Germany, where the coordinated market has built-inrules that help to protect broader social interests. But the finding is surprising for

6

8

10

12

14

16

18

6 8 10 12 14 16 18

industry success

go

vern

me

nt

succ

ess

Britain

Germany

US

France

22

Britain, where the liberal market economy might be expected to underminecollective social goals. France offers another surprise, as its tradition of stateintervention in industry has done little to help achieve the government’sperformance goals. Finally, in the United States, where the Food and DrugAdministration has exerted extraordinary regulatory control over product qualityin the pharmaceutical industry, government goals have similarly not been served.

Before proceeding to an explanation of these differences, I want toaddress a common set of arguments about industrial success in thepharmaceutical sector that focuses on drug testing and intellectual property.Given the strong differences in performance of pharmaceuticals in France,Germany, Britain, and the United States, we see a surprising uniformity in theregulation of testing and intellectual property.

Figure 17. Average effective patent life of 100 main drugs by country.

1980 1988

France 13.6 10.2

Gemany 11.8 10.5

Britain 13.1 11.9

US 15.1 10.7Source: Tarabusi and Vickery 1996, 84.

First, and despite extended debates on national differences in thetreatment of intellectual property (Tarabusi 1996, 129), all four countries nowendorse a 20 year patent protection for new drugs. For France and Germany thisstandard was established by the European Patent Convention, which created aunified patent system in 1972 (OECD 1985, 44). In Britain, patent life wasextended in 1986 from 16 to 20 years (Earl-Slater and Bradley 1996, 397).Moreover, the drug evaluation process, which takes longer today than in thepast, appears nonetheless to take roughly the same amount of time in each ofthe countries. See Figure 17 above. France has the shortest effective patent life,at just over ten years, while Britain has the longest, at just under twelve years.Some countries have offered the possibility of patent extension. In the UnitedStates, the Waxman-Hatch Act of 1984 has permitted patent extensions for up to5 years in case of regulatory delays so as to provide a total effective protection ofup to 14 years (Scherer 1993, 100). France legislated a similar provision for 5year patent extensions in 1991 (Hancher 1992, 392). It remains unclear,however, what an optimal patent life might be for pharmaceutical products, and inany case the average effective patent life appears not to vary strongly among thefour countries in this study

Second, the process of clinical testing has been nearly standardizedacross these countries. In Europe, the results of clinical tests have beenaccepted by member countries since 1977 (OECD 1985, 39). And despiteconsiderable debate about the time required to apply for marketing licenses, a

23

recent study by the European Commission (1997, 31) shows that licenseapplications in France, Germany, Britain, and the US all required an average ofapproximately 2 years. Finally, standards of drug manufacturing across thesecountries are, while not identical, very similar. France, Germany, and Britain relyon World Health Organization (WHO) manufacturing guidelines for drugs. TheUnited States relies on FDA manufacturing guidelines, which are essentially thesame as the WHO guidelines (Balance et al. 1992, 142).

If the regulations that govern testing and intellectual property are similar ineach of the countries studied, strong differences persist in regulation that focuseson the qualities of the products themselves and the markets into which they aresold. Despite ongoing efforts at the European level to standardize thesedivergent product market regulations,10 nonetheless “…existing member stateregulations and concerns regarding the safety and efficacy of pharmaceuticalshave consistently thwarted legislation focused on removing barriers in thepharmaceutical market” (Smith 1995, 471-472). In the following section I explorethe historical development of distinctive national product market regulationsgoverning pharmaceutical products and the impact that these regulations havehad on production strategies.

III. Government Regulation of Drug Quality and Price

Pharmaceutical product regulation in the advanced industrial countrieshas proceeded in two waves. The first wave, stimulated primarily by theThalidomide (Contergan) tragedies, began in the 1960s and focused on drugsafety, efficacy, and testing. The second wave of regulation emerged mainly inthe late 1980s around pharmaceutical pricing and the rapidly increasing cost ofdrugs. The way in which different countries have addressed these issues hasdepended in turn upon national traditions of competition and business regulation,as well as on the organization and political strength of pharmaceutical producersand drug consumers. These distinctive national political and institutionaltraditions have shaped the way in which domestic markets for drugs have beenregulated in each country. Differences in market regulation have, in turn, had animpact on the product strategies of domestic producers.

Drug safety and efficacy regulation has in most countries proceeded intwo steps. The first step has generally required all drug producers to register theirproducts with an administering authority. The second step has required a positivemarketing license for product sale. By the early 1970s most countries had movedfrom passive drug registration scheme to mandatory drug licensing throughwhich all new drugs were required to pass. Within this general trend, however,there has been a great deal of room for national variation in treatment. In France,safety and efficacy oversight by the government has long depended upon theexpertise and information of industry. The result has been a system of industry

10 The European Union has endorsed three approaches to pharmaceutical certification: existingnational certification, a new centralized European certification, and a decentralized multi-stateapproach that requires that producers gain certification in at least two member states. Mostmanufacturers continue to use existing national certification procedures (Juès 1998, 73).

24

capture that has granted the French pharmaceutical industry a broad leeway todetermine the level of efficacy of their products. At the opposite extreme, theUnited States in the 1962 Keffauver-Harris amendments to the Food and DrugAct imposed extremely stringent safety, quality, and efficacy requirements ondrug manufacturers, and they have shown extraordinary resistance to industrylobbying. They have created the most stringent drug standards of any country.

If France and the United States sit at the extremes of drug qualityenforcement, Germany and Britain have taken up an intermediate position. Bothwere relatively late to regulate drug safety and efficacy, and both have reliedheavily on industry to police itself. In Britain, safety measures were imposed bythe 1968 Medicine Act, as well as through an industry code of conduct negotiatedbetween the Association of the British Pharmaceutical Industry (ABPI) and theHealth Ministry. The German pharmaceutical industry has been politicallypowerful and has managed to create for itself a favorable regulatory climate.Germany first required positive drug certification in 1976, but administrativeoversight is less exacting than in the United States and mandatory efficacythresholds are correspondingly lower.

A second wave of pharmaceutical regulation has focused on reducing theoverall drug bill in each country. For some countries, especially Germany and theUnited States, which have traditionally had no drug price regulation, thisrepresents an entirely new demand condition for industry. For France and Britain,recent reforms come in the wake of a long tradition of price regulation that hasshaped pharmaceutical demand for most of the postwar period. From theVarieties of Capitalism perspective this pairing looks unfamiliar, grouping as itdoes France with Britain on one side and Germany with the United States on theother. However, it follows directly from the organization of health care in the twocountries. In France and Britain, the government is in effect a drug monopsonist.In Germany and even more in the United States, a plurality of private healthinsurance companies share the national drug bill.

Two broad strategies have been adopted for restraining the domestic costof pharmaceuticals. The first focuses on restraining the price level of individualdrugs. In this approach, pharmaceutical firms are seen as monopoly producersdriven by profit concerns toward higher pricing. The second strategy for reducingthe domestic cost of pharmaceuticals has focused on lowering overall demandfor drugs in the health marketplace. This approach includes incentives forconsumers to purchase fewer drugs, incentives for doctors to prescribe fewerdrugs, and a limitation on the kinds of drugs that can be reimbursed throughhealth insurance. Each of the countries in this study has resorted to somecombination of these strategies. The historical trend of per capita spending onpharmaceutical products (see Figure 8 on page 13) shows that Britain hasconsistently enjoyed the greatest purchasing restraint. The French experienceshows that price controls alone may fail to restrain overall domestic per capitadrug spending.

France and Britain represent cases in which price regulation by thegovernment has been intensive for most of the postwar period. In Britain,however, industry has enjoyed a strong bargaining position in price-setting

25

negotiations, and the government has acceded to high levels of return on capitalin exchange for a combination of export emphasis and a moderate domestic drugbill. Largely as a consequence of this approach, British pharmaceutical firms areamong the most profitable in the world. In France, by contrast, pharmaceuticalproducers have exercised little control over drug pricing. This had keptpharmaceutical profits low. It has also pushed French producers to emphasize astrategy of low cost production, low innovation, and high volume sales. Bycontrast, Germany and the United States represent cases of minimal governmentintervention in pharmaceutical price setting, at least up until the past decade.

National approaches to pharmaceutical product market regulation haveemerged at the intersection of distinctive national institutions of production(liberal versus coordinated) and of distinctive institutions of consumption (singlebuyer versus multiple buyer). See Figure 18 below. National productioninstitutions have a strong impact on national approaches to the regulation of drugquality. In coordinated economies, product quality is assumed to emerge fromthe highly regulated institutions of production. Thus France and Germany havenot enacted very strict drug efficacy requirements. But national consumptioninstitutions also play a central role. In countries with multiple purchasers, such asthe United States and Germany, the government faces little pressure to enactstrict price regulation. Taking the impact of production and consumptioninstitutions together, we observe in France for example a strategy of strict priceregulation combined with industry dominance in setting drug quality. In the UnitedStates, by contrast, we observe a strategy of strict quality regulation combinedwith a liberal set of price provisions.

consumption

many buyers single buyer

liberal

production

US(quality regulation)

Britain(regulation vianegotiation)

coordinatedGermany

(regulatory laxity)France

(price regulation)

Figure 18. National institutions of drug production and consumption,and the resulting regulatory strategy (in parentheses).

26

Interestingly, the deployment of these regulatory strategies has often ledto unexpected or contradictory outcomes. France’s strict price regulations, forexample, have caused industry to pursue a high-volume, low-margin salesstrategy that has generated an extremely high per capita drug bill. In the UnitedStates, stringent product efficacy requirements had the unintended effect ofdriving many producers out of the country in order to compete in less regulatedforeign markets. Pharmaceutical companies have proved particularly savvy inlocating production so as to take advantage of differences in national regulatoryapproaches (Braithwaite 1993, 16).

In Germany, where the institutions of the coordinated market economyconfront multiple demand-side actors, pharmaceutical product regulation hastended to be lax. The assumption is that a skilled labor force and long-termcapital create their own pressures towards quality production. Moreover,Germany’s multiple medical insurers (Kassen) have not until recently been ableto coordinate their efforts in order to achieve price restraint. In Britain, bycontrast, the combination of a liberal production regime and a unified buyer hascreated concern about both quality and price. The result has been a set ofunusually-broad negotiations among pharmaceutical producers and thegovernment that has generated a regulatory strategy that meets the goals ofboth.

This account of national regulatory traditions is of course stylized. Eachcountry has a distinctive history and regulatory tradition that has conditioned itsresponse to the dual threats of drug safety and drug price. In the remainder ofthis section I lay out in more detail the historical evolution of quality and priceregulation of the pharmaceutical industry in these countries. Each countrydescription presents first safety and efficacy regulation, then price regulation.

Germany: Regulatory Laxity

Pharmaceutical regulation in Germany has intervened in only a minimalway to set safety and efficacy requirements for drugs. Regulatory interventionbegan in 1961 in response to the Contergan (Thalidomide) tragedies. The firstDrug Law (Arzneimittelgesetz, or AMG) of 1961 required only that all drugs incirculation be registered. This weak requirement was reinforced in May 1976 by anew AMG, instigated by the Social Democratic-Liberal Democratic coalition, thatimposed a system of mandatory certification for all drugs (Borchert 1983, 195).All drugs had to be certified by the Bundesgesundheitsamt (BGA) in Berlin, notmerely registered as under the previous law, although companies were givenuntil 1990 to have all of their drugs registered. The staff and funding of the BGAwere increased to take on the extra responsibility; staff grew from 116 to 138,funding grew by 22m DM. The new AGM also called for mandatory labeling ofdrugs (including instructions for use and side-effects), restrictions on drugadvertising, and the requirement that patients give consent to all clinical drugtests. Furthermore, the law imposed a strict standard of product liability on drugcompanies, and required that they purchase liability insurance from a privateinsurer of their choice (Arzneimittelrecht 1976, 147). Insurers were grouped into

27

the Pharma-Pool (Jaeger and Kaukewitsch 1998, 9). Section 84 of the 1976AMG is the first case in which strict product liability was applied in Germany. Butdamages are limited to 500,000 DM for single cases and to a total of 200 millionDM for total damages resulting from a single drug (Rambow 1977, 163).Germany has resisted a formal system for reporting the harm caused by drugs(Test 1983, 2).

The new AMG, which remains the foundation of drug safety in Germany,represented a clear victory of the SPD, which favored a regulatory strategybased on administrative regulation. The CDU had sided with the peak employers’association (BDA) in favoring industry self-regulation. Instead of governmentcertification, they had proposed a collaboration between drug manufacturers andthe medical profession to establish a system of drug labeling that would help toincrease market transparency. The SPD position also did not reflect the view oftrade unions, which, as expressed by the peak German trade union association(DGB), favored a collaboration among drug companies, doctors, and insurers toevaluate the price-performance, quality and usefulness of drugs and to setstandards for how that information should be included on the drug packaging(Bruck 1974, 136-139). One impact of the regulatory solution, however, was toincrease the overall quality and usefulness of drugs in Germany, and thepharmaceutical industry appears to have felt that this would improve themarketability of their drugs overseas (Schatz 1983, 386). Moreover, as theregistration requirement did not have to be met for all drugs until 1990, the costof transition was spread over a period of ten year.

Germany’s relatively late move to regulate drug quality was mirrored by alate effort at restraining drug costs. Until 1989, German pharmaceutical prices aswell as pharmaceutical demand conditions were largely unregulated. Since themid-1970s, the German government had attempted to restrain the cost ofpharmaceuticals. Germany's 1977 Federal Law on Cost Containment establisheda Concerted Action Committee (KAG) to set spending ceilings on the SicknessFunds (Kassen) for drugs paid for by the National Health Insurance Fund (GKV),and granted a joint committee of doctors and Kassen the right to make anegative list of drugs to be excluded from reimbursement (Hancher 1989, 90).Moreover, the 1976 Medicines Act created a Transparency Commission ofindustry, Kassen, and health professional representatives, to publish'transparency lists' comparing drugs based on safety, efficacy, and treatmentcost. Questions of efficacy, however, fell exclusively under the 1976 law to theHealth Department (Bundesgesundheitsamt). Thus industry was able to objectsuccessfully to the use of many comparative criteria by the TransparencyCommission. In 1985 the Supreme Court found that publishing a transparencylist infringed on the right of entrepreneurial activity, and the negative list idea wasset aside. Only an amendment to the Medicines Act of 1986 rendered the use ofa negative list legal (Hancher 1989, 90-91).

In the wake of this change, and under growing cost pressure, the HealthCare Reform Act of 1988 put in place a combination of four separate regulatorymechanisms to manage the drug bill in Germany. A first provision, referencepricing, uses drug reimbursement levels to control the level of drug pricing. Three

28

other mechanisms focus not on price but instead on restraining the volume ofdemand for drugs. A black list eliminates certain drugs altogether fromreimbursement. Indicative volume controls set target spending levels for doctors.Mandatory prescription co-payments also give consuming patients an incentiveto use fewer drugs. Together these four mechanisms constitute a formidable newregulatory nexus that may seem surprising in an age of deregulation and in acountry for which market competition has been a dominant ideology.

The first mechanism, reference pricing (Festbeträge), sets a single level ofreimbursement for all drugs that fall within common pools defined in terms ofspecific common therapeutic qualities. These price levels are set by theorganization of doctors and medical insurers after discussion with industry andpharmacists. Producers are not bound to this price, but patients are required topay any amount by which drugs exceed the reference price for their pool. Inpractice manufacturers rarely exceed these reference prices. Because thesegovernment price levels have encouraged oligopolistic collusion, they haveworked to eliminate price competition below the reference price (Danzon 1997,19-20). Generic drug sales in Germany, for example, remain the lowest of all ofthe countries studies. Only 30 percent of reimbursed drugs fall within thereference pricing system, with the result that some firms have simply stoppeddeveloping new drugs in these therapeutic categories (Finkelstein and Bittinger1993, 18). To remedy this problem, the government in 1993 imposed a pricefreeze on all drugs not covered by reference pricing, and set the frozen level at95% of the March 1992 level (Jaeger and Kaukewitsch 1998, 11). Thus allprescription drugs were covered either by reference pricing or by the pricefreeze. By 1997, 64 percent of all prescription drugs were covered by referencepricing.

In addition to price controls, three other regulatory mechanisms haveworked to hold down demand for drugs. The first of these demand-managementefforts was a negative, or black, list that eliminates certain drugs thought to be of“questionable therapeutic value” from reimbursement. This list is drawn up by theMinistry of Labor (Jaeger and Kaukewitch 1998, 10). About 2,500 drugs havebeen blacklisted in this way. But this is a relatively small part of the total market,and leaves an additional 13,000 kinds of drugs that are still available forreimbursement (Finkelstein and Bittinger 1993, 17). For the pharmaceuticalindustry this black list has the effect of concentrating competition in coretherapeutic categories that have in any case been the focus of traditionalGerman success.

The 1988 Health Care Reform Act also imposed a set of volume controls(Richtgrössen) on physicians that set a recommended drug budget for differentcategories of medical practice. The system is monitored by physicianassociations. Doctors who exceed their recommended level facing an audit(Wertheimer et al. 1996, 166). To compliment these volume controls, intended toreduce physician demand, Germany also imposed in 1994 a co-payment feeintended to made patients more conscious of the cost of drugs they consumed.The co-payment was set first at 3.5 DM or 7 DM, depending on the price of thedrug. In 1997 the co-payment was raised to 9 DM, 11 DM, or 13 DM, depending

29

again on the expense of the drug. Pharmacists strongly opposed this measure,and indeed experienced a 10 percent drop in profit when the co-payment systemwas put in place (Jaeger and Kaukewitsch 1998, 11).

What impact have these regulations had on the pharmaceutical industry?The final effect will be unclear for some time, but initially at least they havereceived support from industry for the benefits that they convey. The referencepricing system has helped the industry to create oligopolistic prices that have,among other things, repressed the market for generic drugs (Danzon 1997, 19-20). It has also forestalled a significant threat from inexpensive French drugs.The black list helps to set a high standard of competition, by eliminatingreimbursement for more frivolous kinds of drugs, even if they have passedregulatory safety and efficacy requirements.

The system of volume controls for doctors also represents a mild victoryfor the drug industry. First, it emphasizes demand rather than price restraint.Second, the burden of restraint falls on the medical profession rather than on thepharmaceutical industry itself. Moreover, by forcing doctors, who are arguablymost capable of deciding rationally about drug consumption, to make harddecisions about where limited drug spending must go, the regulation creates theconditions for a strong competition focused on quality and usefulness. In anycase, the impact of domestic regulations on industry in Germany is likely to beless than in other countries, as only 37 percent of sales serve the domesticmarket (Finkelstein and Bittinger 1993, 18).

Britain: Regulation via Negotiation

Drug quality regulation in Britain is relatively rigorous, although Britishdrug reviews tend to emphasize safety more than efficiency. Basic safety andefficacy requirements were set first by the 1969 and 1971 British Medicines Acts.As in other European countries, Britain requires that drugs receive a marketlicense certifying their safety, efficacy and quality (Earl-Slater 1996, 396). Theseevaluations are sub-contracted to expert committees. Britain has been moreaccepting of foreign trials than have other countries (Grabowski 1981, 9). ButBritain also relies on export and research incentives to induce pharmaceuticalcompanies to self-police the safety of their products. And it relies more heavilythan other countries on a post-marketing evaluation of how drugs are performingin everyday use (Tarabusi 1993, 129).

Most importantly, the way in which pharmaceutical pricing is set in Britaincreates incentives to pursue safe and innovative drugs. Britain, like France, has along tradition of government regulation of pharmaceutical pricing. This high levelof regulation is perhaps surprising in the VOC framework, but it reflects the realitythat the British National Health Service pays for 90 percent of all drugspurchased, the highest government share of any of the countries studied. Since1958, British pharmaceutical manufacturers have negotiated directly with thegovernment in order to set the sales price of drugs. But the British system doesnot focus either on individual product prices, nor on company revenues, butinstead on return on capital. Britain is the only country in Europe to focus controlson pharmaceutical profits (Commission 1997, 70). This approach requires an

30

extremely high level of government oversight, and intervention in areas of basicproducer strategy. But it offers the advantage of permitting British firms to pursuemore risky products that could also be more lucrative (Danzon 1997, 53). Morerecently, Britain has also begun regulating the demand conditions ofpharmaceutical production.

The history of postwar British pharmaceutical regulation can usefully bedivided into two periods. From 1958 to 1978, the pharmaceutical industrynegotiated with the government over voluntary price restraint. Until 1960,negotiations were carried out by the Association of British PharmaceuticalIndustries (ABPI). After that the government negotiated directly with many of thelarger producers. This approach was called the Voluntary Price RegulationScheme (VPRS). By 1969 prices were based explicitly on return on capital fordrugs purchased by the National Health Service (Sargent 1985, 112-117)

A second period, beginning in 1977, was marked by the end of voluntarynegotiations. The election of the Labour Party in 1974 carried with it the threat ofnationalization, which put the pharmaceutical industry in a difficult negotiatingposition. The shift was made decisive when the 1977 Health Act granted theSecretary of State for Social Services the right to impose drug prices(Greenwood 1991, 340). While Britain has never fallen back on this provision, theexplicit right to intervene changed the nature of government-industrynegotiations. Indeed, in recognition of the non-voluntary character of lateragreements, the new price setting scheme was renamed the PharmaceuticalPrice Regulation Scheme (PPRS). The PPRS lies at the core of British drugpricing today.

Under the PPRS, companies are compensated for higher investments andexport success (Sargent 1985, 120-121). Agreements are negotiated annuallybetween the Medicines Control Agency (MCA) of the Department of Health andindividual manufacturers (Earl-Slater 1996, 398). The price that the NationalHealth Service (NHS) pays for drug products is set in these negotiations, basedon the return on capital employed (ROCE) for those drugs that the NHS buys.(OECD 1985, p 40). In the 1993-98 period, allowed ROCE was between 17 and21%. These levels are not guaranteed, but producers are permitted to increasedrug prices in order to achieve their negotiated range (Earl-Slater 1996, 399). Inthe case of excessive profits, producers are required to offer rebates directly tothe government (Sargent 1985, 122). What level of ROCE a producer may claimdepends in turn upon the level of their investment in research and development,and upon the export success of their products. This means that the profit of anyBritish pharmaceutical company is determined almost entirely by their innovationand upon their level of export.

Apart from controls on the profit of pharmaceutical manufacturers, Britainhas also implemented four projects to lower the demand for pharmaceuticals.Many of these are similar to German efforts to keep down demand for drugs.First, in 1985, the British government published a black list of 600 drugs not to beprescribed by NHS doctors. Beginning in 1993 the list was extended (Earl-Slater1996, 400-401). Second, since April 1991, doctors have been given indicativedrug budgets. Prescription levels of different categories of doctors are compared

31

through the nation-wide computerized Prescribing Analysis and Costs (PACT)system. As in Germany, doctors who over-prescribe are audited, in Britain by theHealth Commissioner. Starting in 1991, in an effort to introduce marketmechanism among prescribing doctors, some general practitioners have become“fund holders,” able to make their own decisions about dispensing a pool of fundsfor their patients. Third, Britain has long imposed patient co-payments onpharmaceutical prescriptions. These have been increased from 0.45 pounds in1979 to 5.50 pounds in 1996. However 50 percent of the population is exemptedfrom these fines, accounting for 80 percent of drug use (Earl-Slater, 1996, 403-405).

Finally, beginning under the Labour government of 1974-1979, a 14%ceiling was set on promotional activities by the drug industry. This level droppedto 9% in 1983-1984. About half of this promotional spending goes to payingmedical representatives who visit individual practitioners to push specificproducts (Greenwood 1991, 344-345). Indeed the high percentage of sales spenton R&D in Britain is due in part to the relatively low promotional spendingmandated by national regulations (Finkelstein 1993, 5). But because thegovernment has kept permissible profit margins relatively high (between 18 and21 percent or revenue), and because these profit margins are not related todomestic sales levels, the drug industry has accepted lower promotionalspending.

Given the profound influence of Britain’s Pharmaceutical Price RegulationScheme (PPRS) on industry profit, innovation, and sales, it seems unlikely thatBritish pharmaceutical success can be attributed solely to the benefits of a liberalmarket economy. Contrary to the notion of a liberal market economy, industryand the government have long negotiated a set of industry targets that weremutually acceptable. This does not mean that the VOC model was irrelevant. Thegoals of industry — high profit and high investment in R&D — were exactly thoseenvisioned by the VOC model. Yet the means by which they were achievedsuggests that industry success may depend as much on the configuration ofconsumption as on the configuration of production.

France: Price Regulation

The French regulatory approach to the pharmaceutical industry is unusualin that it emerged directly from the regulation of apothecaries (Plat 1970, 7). Ithas been characterized by a long history of industry self-regulation of productquality combined with government oversight of product pricing. Drug safety andefficacy evaluations have, until very recently, been carried out by industry expertspaid for by pharmaceutical companies and employing data from the applyingfirm. Patricia Danzon estimates that only half of all drugs sold in France wouldmeet US Food and Drug Administration efficacy requirements (Danzon 1997,44). Drug pricing, by contrast, has been the focus of intensive governmentintervention since the 1950s.

Drug quality regulation in France began under the Vichy government in1942, with the creation of the Service Central de la Pharmacie. Its primaryemphasis was not on drug safety but instead on ensuring the quality and novelty

32

of drug production, much of which occurred at individual pharmacies (Hancher1990, 107; Buisson and Giorgi 1997, 20). Drug producers applied for a visa thatwas granted based on the recommendation of experts from the industry. The visarequired that any drug be both useful, and also new (Brudon and Viala 1995,215). This novelty requirement permitted the visa system to function as a kind ofpatent system before patents were applied to drugs. Under the visa system,newly registered drugs were granted a 6 year window of monopoly sales. In 1959a special patent was introduced for medicines and the novelty requirement wasremoved from the visa application process (Buisson and Giorgi 1997, 19). Thevisa system also required a manufacturing authorization (autorization de débit)from the French pharmaceutical inspectorate. This requirement effectivelyblocked the import of finished drugs into France (Hancher 1990, 115).

In the wake of the Thalidomide disaster, and under the guidance of EECPharmaceutical Directive 65/65, France enacted the ordinance of 1967 whichreplaced the visa system with a new marketing authorization (autorization demise sur le marché, or AMM) and drug test protocols. The system was overseenby a new Commission on Marketing Authorizations (Commission d'autorisationde mise sur le marché des medicaments) with 28 members. These wereprimarily doctors, but also included two representatives from industry and oneconsumer representative (Hancher 1990, 124-125). As under the visa system,the expert remained central to the new approach. Experts were paid by thecompany requesting an evaluation. Moreover all information about the drug camefrom the company itself (Hancher 1990, 126-127).

In 1977 the Comission of Marketing Authorizations was replaced again bythe Direction de la pharmacie et du medicament (DPHM) of the Health Ministry(Buisson and Giorgi 1997, 20-21). This new centralized body was responsible forall aspects of drug regulation, but had a staff of only 30! Hence, as before, drugapplications were assessed out of house by experts paid by the applying firm(Hancher 1990, 124). In 1978, France imposed a set of guidelines for goodmanufacturing practice (bonnes pratiques de fabrication, or BPF) that drew on acommon European standard (Buisson and Giorgi 1997, 26). In addition to theformal approval process for drugs, the Ministry of family health in 1976 created aCommission technique de pharmaco-vigilance. Its mission was to collect andcompile information about drug-related accidents. It's members were eminentprofessors. Consumer groups and the Commission technique de pharmaco-vigilance worked together in the Liaison avec les organisations deconsommateurs (Pujol 1978).

In 4 January 1993, the DPHM was restructured and renamed the Agencefrancais du medicament (AFM). The new agency was given responsibility forsetting safety standards, for setting manufacturing policy, for deciding on productinformation and risk, and for assisting industry to further research and innovation.Unlike the previous DPHM, the AFM has its own internal experts that evaluateindustry applications for AMM (Agrément de Mise sur le Marché). The AFM setsthe labeling and instructions for new drugs. The AFM also sets the medicalclassification of the drug, i.e., prescription of over-the-counter (Buisson andGiorgi 1997, 20-21). Like the US FDA, the Agence Francais du medicament

33

enjoys a high level of insulation from political interference, although there isprovision for an emergency procedure in case of a “serious menace to the publichealth” (Viala and Vion 1994, 246-248).

In the regulation of drug pricing, France combines the lowest prices forindividual drugs with the highest per capita spending on drug purchases. Thispattern has arisen from a set of regulatory policies that have effectivelyrestrained drug prices but have been much less effective at holding downdemand. Price regulation has been the primary means of restraining monopolypricing in the drug industry for most of the postwar period. Only in the past tenyears have French regulators acknowledged that these price restraints have hada negative impact on the French pharmaceutical industry, at least from theperspective of the government. First, they have failed to keep down the overalldrug bill in France compared to other countries. Second, they have kept industryinvestment in research and development low. These are problems for thegovernment but not necessarily for the pharmaceutical industry itself. Operatingin what Hancher (1990, 229-230) has called a guilded cage, French drugproducers have responded by focusing their production efforts on a strategy oflow price and low innovation combined with high promotional spending.

Price fixing is a revered tradition of French product markets, dating back atleast to the ordinances of 1945 (45-1483, 45-1484) which granted thegovernment the right to intervene to set prices for any kind of product. Pricesetting in the pharmaceutical industry began in 1948 and can be usefullyseparated into four periods. Between 1948 and 1968, drug prices were set by thestate based on a pricing framework (cadre de prix) calculated on a cost-plusbasis from a combination of production costs and a pre-determined margin ofprofit (Hancher 1990, 76). In 1968, drug pricing fell under the dual control of theold pricing administration and the new Commission de reimboursement de lasecurite sociale (the Coudrier Commission) created in 1967. While drug priceswere still set by the state pricing administration, the new social securitycommission worked with them to determine reimbursement levels for prescriptiondrugs, and to decide which drugs should be included on the reimbursement list(Hancher 1990, 92-3).

In 1980, drug prices for prescription drugs were liberalized. This meant inpractice that drug prices would be determined entirely by reimbursement levelsset by the Social Security code. In order to decide an appropriate reimbursementlevel a new Commission de transparence, created in 1980, evaluated priceapplications from drug producers. They assessed production costs andcompared applicant drugs with others on the market (Juès 1998, 70). Under thissystem, a separate price was set by the government for each individual drug.The 1986 Plan Seguin to reduce the cost of health care included among its cost-cutting efforts a reduction in state reimbursement for drugs and a cap on drugprice increases to about 1 percent per year.

Concerned that this approach reduced industry incentives to innovate, andin particular that it was leading French drug companies to create false-innovations in order to elude the cap on price increases for existing drugs,France created in 1994 a pricing system based on a so-called global envelope

34

(envelope globale). The idea, first proposed in 1991, allowed the state tonegotiate with each pharmaceutical producer an upper and lower level for theirexpected sales level for the year. Any sales exceeding the target level wouldhave to be offset by a lower unit price on drugs so that the overall cost to thegovernment would remain the same. The goal was to give producers the freedomto set prices as they wished among their portfolio of products. So as not toimpede major innovations, breakthrough drugs were treated separately under theold price-setting scheme (Finkelstein 1993, 12-3; Danzon 1997, 18-19). By mid-1995, 96 conventions had been signed under the envelope globale system,covering 81 percent of drug sales (Hofstetter 1996, 15).

While France’s primary effort at reducing medical spending was focusedon price control, it has also more recently attempted to create incentives on thedemand side in order to keep down spending. Since 1980, drugs have beenreimbursed under social security at different levels depending on effectivenessand need. Reimbursement is set at 100 percent for cases of severe sickness, 65percent for regular drugs, and 35 percent for drugs with “modest therapeuticvalue” (Juès 1998, 77). Most patients, however, have a private insurance planthat covers the non-reimbursed cost of drugs. Only 10 percent of French patientsare not covered in this way (Wertheimer 1996, 162-3). More recent efforts havefocused on restraining doctor prescription levels. A 1990 controle medicale setnon-binding, indicative levels for doctors in order to keep them from over-prescribing (Finkelstein 1993, 12). In January 1994 a new set of negotiated goodmedical practice guidelines has attempted to standardize prescription activities ofdoctors. (Wertheimer 1996, 162-3) The impact of the system, also voluntary,remains uncertain.

France has also acted to reduce the level of drug advertising. As in theother European countries, advertising to the public for drugs covered by socialsecurity is prohibited in France (Juès 1998, 70). Nearly 80 percent of promotionalspending by French firms has therefore focused on informational visits to medicalpractitioners (SNIP 1996). As part of the 1994 framework agreement thatimplemented the global envelope pricing system, the peak association of thedrug industry, SNIP, agreed to voluntary restraint on advertising spending inexchange for more liberal prices for innovative products (“L’Industriepharmaceutique” 1995, 4; Danzon 1997, 18-19). French firms nonetheless spendalmost 13 percent of revenue on advertising and marketing, compared to onlyabout 7 percent on research and development (SNIP 1996).

35

Peculiarities of the French pricing scheme have pushed companies toevade the effects of regulation. Two side-effects were particularly formative forthe drug industry. First, research costs were fixed at 7%, so that in effectcompanies that did less research enjoyed higher profits. Because the prices ofexisting products were not systematically reviewed, producers commonlyreintroduced existing drugs under new names in order to receive a higher price.Useful inexpensive drugs disappeared from the market and were replaced by alarge number of copy drugs (Hancher 1990, 89-90). Alan Afuad (1993, 9-10)writes, “price controls contribute to an increase in drug expenditure…becauseresearch and development is re-oriented towards developing non-innovative newproducts with contemporary price tags to replace older drugs with low controlledprices.” Second, because prices were calculated based on production costs,many French firms moved component production into their own subsidiarycompanies located in foreign countries. Through exclusive contracting with thesesubsidiaries, the firms could artificially elevate the prices of imported chemicalcomponents while extracting higher profits from their foreign firms (Faibis 1984,23; Hancher 1990, 94).

50%

60%

70%

80%

90%

100%

110%

120%

130%

140%

150%

1965 1968 1971 1974 1983 1988 1993 1997

Germany

France

Britain

USA

Sources: EC 1997, p. 71; Faibis 1984, p. 22; Möbius et al. 1976, p. 75; SNIP.

Figure 19. Index of relative drug prices in France, Germany, and Britain,1965-1997.

36

French price restraints have indeed been effective at keeping down theunit cost of drugs. See Figure 20 above. French firms have adjusted by skimpingon R&D and advertising their products heavily. The combination of ineffectiveand inexpensive drugs, pushed by a strong advertising campaign, has created asurge in consumer demand. This is why France combines the lowest unit drugprices with the highest per capita drug spending. Furthermore, these low prices,publicized by the government, have made French pharmaceutical firms quitesuccessful in exporting conventional kinds of medicines. Low prices have madeFrance a particularly strong exporter of antibiotics and vaccines, for example,especially to developing countries. In sum, a regulatory practice that has been amajor failure from the perspective of the French government, has, by contrast,carved out a low-innovation business niche for French pharmaceuticalmanufacturers.

United States: Quality Regulation

Drug safety regulation in the United States has proceeded in two stages.A first stage, beginning with the Food and Drug Act of 1938, was characterizedby a high degree of industry control over drug markets. A second period of strictgovernment oversight, initiated under the 1962 Keffauver-Harris DrugAmendments, continues to characterize drug regulation today. The United Statestoday is generally considered to impose the highest drug quality and efficacystandards of the advanced industrialized countries. After a decade of strongopposition to this policy, the pharmaceutical industry has come to accept theprinciple of strict regulatory oversight by the FDA as a source of comparativeadvantage for the US industry, as well as an impediment to foreign imports.

The FDA, created in 1906 in response to Upton Sinclair’s fictional accountof the Chicago meat packing industry in The Jungle, gained significant regulatorycontrol over drugs only in 1938, following the death of 107 people fromconsuming an elixir of sulphanilamide (Davis 1997, 15). The 1938 Drug Actimposed drug labeling standards, or alternatively permitted drug companies toevade the labeling standards if they sold their drugs by prescription only. Underthis early set of regulations, producers were required to register new drugs withthe FDA, after which the FDA had ninety days in which to raise objections tocommercializing the drug. In the absence of FDA notification, the company couldproceed with marketing the drug. The process of drug testing was largelyunregulated. Amendments to the Drug Act in 1951 instituted a mandatoryprescription-only class of drugs (Bogner 1996, 76).

This move set the stage for an extraordinary consolidation of power bypharmaceutical manufacturers in the United States. The primary focus of thiseffort was on eliminating the generic drug industry, which amounted to nearly onethird of the drug market in 1950. The new National Pharmaceutical Council(NPC), created 1954 by the large pharmaceutical manufacturers, proved sosuccessful that generic sales fell to less than 5 percent by 1959. Their strategyinvolved co-opting pharmaceutical service providers, namely drug stores anddoctors.

37

Drug stores in the United States, having received a boost duringprohibition (as the only stores that could legally sell alcohol, and thanks to thesoda fountain), had by the early 1950s become large and powerful retail chains(Mobley 1990). Lively competition among drug stores was exerting strongdownward price pressure on pharmaceutical producers. Through intensive state-level lobbying, the NPC was able to block drug stores in most states fromadvertising prices for prescription drugs. In many states a standard mark-up wasprovided for drug stores, typically 33% of the wholesale price, which helped toreduce competition over drug prices (Bogner 1976, 77-78). Finally, and critically,the NPC succeeded in making it illegal for pharmacies to substitute generic drugsfor branded prescriptions. They managed this by petitioning individual states’Boards of Pharmacy to change the legal definition of drug substitution (Hirsch1975, 333).

The NPC launched a similarly effective campaign to promote brandeddrugs in the medical profession. In what appears to have been a quid pro quoagreement, the American Medical Association (AMA) shifted their support awayfrom generic drugs and towards branded drugs in exchange for greateradvertising revenues in journals published by the AMA (Bogner 1976, 77). TheAMA-administered pharmacopoeia Useful Drugs, in which drugs were evaluatedand categorized by generic name, was superceded by the Physician’s DeskReference, compiled from manufacturer data by the newly createdPharmaceutical Manufacturers Association (Pharma), created in 1958. The AMAbegan encouraging physicians to prescribe branded rather than generic drugs.Second, when the drug industry came under increased government scrutiny inthe 1960s, the AMA argued against requiring a higher level of drug efficacy. AsPaul Hirsch has written, “it is clear that the medical profession was effectively co-opted by the drug industry” (Hirsch 1975, 338-339).

Strict government oversight of pharmaceuticals, which began with the1962 Kefauver-Harris amendments, arose largely as a backlash against thisgrowing power of the drug industry. It also followed the revelation of birth defectscaused by Thalidomide, a drug which the FDA had warned against but was at thetime was unable to block for experimentation. In 1960, Senator Keffauvercharged the pharmaceutical industry with monopolistic exploitation of consumers(Schwartzman 1976, 325). The 1962 amendments required a positive approvalfrom the FDA for a new drug to be marketed, and called on the FDA to establishgood manufacturing practice (Edwards and Glenn Thomas 1983, 58-59). Theyset high standards not only for drug safety but also effectiveness. They alsorequired a preliminary FDA approval before undertaking clinical testing(Grabowski 1981, 7).

This legislation fundamentally changed the industry. The cost of creating anew chemical entity (NCE) rose by a factor of 2 over the decade following itsenactment (Grabowski et al. 1978, 157). Development times grew from 3-4 yearson average for each NCE before 1962, to 7-10 years after that date (Grabowski1981, 8). Moreover, a lag in US pharmaceutical performance in the 1970s wastaken to indicate that the more stringent FDA approval process called for in the1962 amendments had hurt innovation in pharmaceuticals (Grabowski et al.

38

1978, 137). Critics argued that the increased cost of innovation had pushedcompanies to lower their investments levels (Edwards and Glenn Thomas 1983,66-67).

Subsequent congressional opposition to the increased power of the FDAwas extremely vociferous. In the years 1969 and 1970 alone, Congressionalhearings were held on 38 FDA decisions not to approve the marketing of specificdrugs. Congress in other words was putting heavy pressure on the FDA toapprove drugs (Edwards and Glenn Thomas 1983, 62), but the FDA was andremains insulated from such political pressures. Today, strict FDA oversight isseen as an important comparative advantage for US drug manufacturers. Thestandard of safety and efficacy set by the FDA remains unusually high byinternational standards. Japan’s top-selling drug in 1986, for example, an anti-cancer agent called Krestin, has never been submitted for US approval in theunderstanding that it would not be likely to pass the FDA review (Reich 1990,140). This means that drugs originating in the US market meet a particularly highstandard of safety and efficacy and are strong competitors for foreign sales. Italso means that many foreign drugs of questionable effectiveness are kept out ofthe US market.

But this high domestic standard of safety and efficacy, in conjunction withidiosyncratic drug licensing requirements in foreign countries, had theunexpected effect of encouraging US firms to move production overseas. Underthe strict US standards, drugs cold not be exported if they had not yet receivedFDA approval, even if they would be acceptable in the receiving country. UScompanies thus established foreign subsidiaries, operating under foreign safetyand efficacy rules, to produce products for foreign consumption. In this way, thehigh safety and efficacy requirements of the United States exerted a strongpressure for globalization of company ownership, and a corresponding reductionin export of drugs produced in the United States. The US prohibition againstexporting unlicensed drugs was rescinded in the 1986 Drug Export AmendmentAct, but by that time the strategy of overseas production by US firms had beenfirmly established (Tarabusi and Vickery 1996, 101-102). This regulatory side-effect must be considered the primary reason for the continued low exportemphasis of domestic drug producers.

In contrast to the strict treatment of drug quality, the United States hasimposed by far the lowest level of regulatory control of drug pricing and demand.This is due primarily to the low level of government spending in the national drugbill. In 1996, 43 percent of the cost of drugs was paid by the US government, viaMedicare, Medicaid, or through the Veterans Administration. Individual insurancecompanies pay slightly less, at 39 percent, with the remainder paid by consumersthemselves (Wertheimer et al. 1996, 169). The most significant incursion into freepricing has been the 1990 Medicaid Prudent Pharmaceutical Purchasing Act(Pryor Bill, Public Law 101-508), which imposed a rebate on governmentpurchases at least 5 percent below standard volume rebates offered by drugcompanies (Scherer 1993). The bill was imposed to permit the government totake advantage of rebates that drug companies were offering to large privateinsurance companies. One unexpected result has been to lower volume

39

discounts to all large purchasers in order to reduce the level of rebate to thegovernment. Moreover, because the Veterans Administration itself has in thepast benefited from large negotiated rebates, it is unclear whether the Pryor Billhas actually decreased the total government drug bill (Finkelstein and Bittinger1993, 19).

Despite the Pryor Bill, the United States remains, viewed in terms of price,the most deregulated market for pharmaceutical products. The relatively low levelof government participation in drug payment has accompanied relatively higherdrug prices than in other European countries. In 1983, for example, a standardbasket of drugs cost nearly twice as much in the United States as in France (SeeFigure 20 above). These high prices have helped to drive up pharmaceuticalindustry profits. In part because of the high cost of drugs, the Clintonadministration has suggested that the government should take on the cost ofpharmaceuticals for many elderly. This move, which would increase thegovernment share of the drug burden significantly, would also be likely to createpressures for greater price regulation. The French example stands as a warningof the potential negative impact of pharmaceutical price regulations in theabsence of more extensive negotiations between government and industry.

IV. Conclusions

From the Varieties of Capitalism literature, we have derived two generalhypotheses about comparative national industrial performance in thepharmaceutical sector. First, because coordinated market economies such asGermany and France are oriented towards high quality but incrementalinnovation strategies, we should expect poor performance in the innovation-intensive drug sector. Second, because institutional setting confers comparativeadvantage in what is a highly globalized industry, we should expect to seegreater trade success among the liberal market economies. We should alsoexpect to see pharmaceutical producers moving production out of France andGermany and into Britain and the United States.

Evidence from the pharmaceutical sector offers strong confirmation of thefirst hypothesis, especially in areas of employment, profitability, and innovation.Coordinated market economies have promoted a high level of employment in thepharmaceutical industry. This is consistent with restrictive labor laws thatincrease the cost to employers of pursuing layoffs. Hence French and Germanpharmaceutical sectors account for a greater percentage of employment than doBritish and US sectors. Profitability in the drug sector in France and Germanyhas been significantly lower than in Britain and the United States. This finding isconsistent both with rigid labor laws in coordinated market economies and withthe availability of risk-accepting capital markets in liberal market economies.Innovation echoes national patterns of profitability. Consistent with the VOCmodel, France and Germany invest relatively little in new drug developmentcompared to Britain and the United States. Thus in many important aspects ofindustry performance, experience in the pharmaceutical industry confirms theinsights of the VOC model.

40

The second hypothesis generated from the VOC model, the expectationthat liberal market economies should enjoy a comparative internationaladvantage in the innovation-intensive pharmaceutical industry, finds less supportin industry performance indicators. While Britain is the most successful drugexporter, as the VOC model predicts, France and Germany have also enjoyedstrong export success. More surprisingly, the United States has not. The reasonsfor this export pattern have less to do with the organization of domesticproduction than with the regulation of domestic product markets. French firmshave enjoyed strong exports because stringent government price regulation hasmade them price-competitive in low-technology areas of pharmaceuticalproduction such as vaccines and antibiotics. US firms have shown weak exportperformance, primarily because they have decided to locate production forforeign markets in foreign countries. Even in Britain, where exports are expectedto be high, pharmaceutical companies have been encouraged to export bymeans of drug pricing regulations that reward greater export intensity with higherprices.

Moreover, patterns of company ownership appear directly to contradictVOC expectations about comparative institutional advantage. The United Statesand Britain, which should enjoy the most salubrious production environment,engage in the most overseas production. France and Germany, which shouldhave the greatest interest in moving product development and production toBritain and the United States, do not. Why? France and Germany may berestrained by strict labor market rules that keep French and German companiesfrom firing locally and hiring globally. The high cost of moving thus makes anexport strategy more attractive for these countries. The United States and Britainappear to have been driven by product market conditions to move productionabroad. Britain’s interest has primarily been to capture a share of the large USmarket. In the United States, however, stringent domestic efficacy requirementsimposed by the 1962 Kefauver-Harris amendments pushed companies to set upoverseas production sites. Although these restrictions have been eased,overseas production remains extremely high. The trend was aggravated bycountries such as France, which until the early 1970s placed an effective ban onimported drugs, thereby encouraging foreign firms to buy domestic productionsites. Finally, the high profit and equity level of British and US firms hasundoubtedly made it easier for them to acquire foreign companies. In sum,patterns of ownership and export have been dictated by a combination ofproduction and product market conditions that have interacted in oftenunexpected ways.

Given the importance of domestic product market regulations to industrystrategy, what kinds of regulations are best? This research suggests that, as inproduction, a coordinated approach to product market regulation generatesstrong benefits both for industry and for society. Britain’s Pharmaceutical PriceRegulation Scheme offers a good example of coordinated regulation, in whichindustry and the government have come to a regulatory solution that respects thefundamental goals of both actors. France and the United States, by contrast,offer examples of uncoordinated product market regulation. In France, where the

41

state has imposed strict price restrictions, companies have responded bypursuing a high-volume, low quality product strategy that has had the unexpectedresult of pushing domestic drug consumption extremely high. In the UnitedStates, where the state has imposed strict efficacy restrictions, companies haveresponded by moving production for foreign markets overseas. In these cases,uncoordinated approaches to product market regulation have hurt performanceboth as it is perceived by the government and as it is perceived by industry. InBritain, where product market regulations have been negotiated, both thegovernment and industry enjoy superior outcomes. This suggests that productmarket regulation does not take the form of a zero-sum trade-off between privatebenefits and public externalities. Instead, a coordinated approach to regulationappears to combine industrial and social benefits.

Beyond these theoretical points, three empirical findings from thepharmaceutical industry should be highlighted. First, the US pharmaceuticalindustry, measured across a broad range of industry performance indicators,ranks close to the bottom. It lags behind Germany in its social benefits andbehind Britain both in social benefits and in industry-relevant performanceindicators. Second, the pharmaceutical industry is a highly globalized industrythat is also highly regulated, with different countries employing very differentregulatory strategies. At least in this industry, global trade and ownershipappears to be compatible with a very high level of domestic regulation. Indeed, arapid increase in drug trade over the course of the 1990s accompanied a rapidincrease in specialized, nationally distinctive drug pricing regulation. Finally, thestrong export success of German pharmaceutical manufacturers, despite theinnovation challenges posed by new technologies, suggests that viable strategiesexist for coordinated market economies even in industries characterized by highrisk and rapid innovation.

42

Reference List

1. L'industrie pharmaceutique. 1995. Economie Géographie 321-322, no. 1-2:1-24.

2. World OTC Retailing. 1995. Retail Monitor International February: 133-45.3. Abel-Smith, B., and P. Grandjeat. 1978. Pharmaceutical Consumption:

Trends in Expenditure, Main Measures Taken and UnderlyingObjectives of Public Intervention in this Field., Vol. 38. Public PolicySeries. Brussels: Commission of the European Communities.

4. Afuad, Alan N. 1993. Public Policy and Pharmaceutical Innovation: ALiterature Review and Critique. Sloan School of Management,Program on the Pharmaceutical Industry, Working Paper 14-96.

5. Albach, Horst. 1994. Culture and Technical Innovation: A Cross-CulturalAnalysis and Policy Recommendations. Berlin: Walter de Gruyter.

6. Albert, Michel. 1993. Capitalism Against Capitalism. Translated by PaulHaviland. London: Whurr.

7. Ballance, Robert, János Pogány, and Helmut Forstner. 1992. The World'sPharmaceutical Industries: An International Perspective onInnovation, Competition and Policy. Hants: Edward Elgar Publishing.

8. Ballance, Robert H. 1996. Market and Industrial Structure. ContestedGround: Public Purpose and Private Interest in the Regulation ofPrescription Drugs. Edited by P. DavisNew York: Oxford UniversityPress.

9. Bogner, William C. 1996. Drugs to Market: Creating Value and Advantage inthe Pharmaceutical Industry~. Oxford: Pergamon.

10. Borchert, Günter. 1983. Artzneimittelgesetz und Verbraucherschutz.Zeitschrift Für Rechtspolitik 8: 194-96.

11. Braithwaite, John. 1993. Transnational Regulation of the PharmaceuticalIndustry. Annals of the AAPSS 525: 12-30.

12. Bruck, Gerhard Wilhelm. 1974. Perspektiven der Sozialpolitik : Synopsesozialpolitischer Vorstellung der Bundesregierung, SPD, FDP, CDU,CSU, DAG, des DGB und der Bundesvereinigung der DeutschenArbeitgeberverbande. Goettingen: Otto Schwartz.

13. Brudon, Jean, and Georges Viala. 1992. France. InternationalPharmaceutical Services: The Drug Industry and Pharmacy Practicein Twenty-Three Major Countries of the World. Edited by Richard N.Spivey, Albert I. Wertheimer, and T. Donald RuckerNew York:Pharmaceutical Products Press.

14. Buisson, Jean-Philippe, and Dominique Giorgi. 1997. La Politique dumédicament. Paris: Montchrestien.

15. Burstall, M. L. 1985. The Community's Pharmaceutical Industry.Luxembourg: Commission of the EC.

16. Burstall, M. L., and B. G. Reuben. 1987. The Cost of Non-Europe in thePharmaceutical Industry. London: Economists Advisory Group.

17. Casper, Steven, and Catherine Matraves. 1997. Corporate Governance andFirm Strategy in the Pharmaceutical Industry. Discussion Paper FS

43

IV 97-20, Wissenschaftszentrum Berlin.18. DAFSA, Eurostaf. 1989. Les laboratoires pharmaceutiques. Paris: Eurostaf

DAFSA.19. Danzon, Patricia M. 1997. Pharmaceutical Price Regulation: National

Policies versus Global Interests. Washington, DC: AmericanEnterprise Institute Press.

20. Davis, Peter. 1997. Managing Medicines: Public Policy and TherapeuticDrugs. Buckingham: Open University Press.

21. Earl-Slater, Alan, and Colin Bradley. 1996. The Inexorable Rise in the UKNHS Drugs Bill: Recent Policies, Future Prospects. PublicAdministration 74: 393-411.

22. Edwards, Charles C., and Lacy Glenn Thomas. 1983. The CompetitiveStatus of the U.S. Pharmaceutical Industry: The Influences ofTechnology in Determining International Industryal CompetitiveAdvantage. Washington, DC: National Academy Press.

23. Elgar, Edward, Robert Ballance, János pogány, and Helmut Forstner. 1992.The World's Pharmaceutical Industries: An International Perspectiveon Innoavtion, Competition and Policy. Paris: UNIDO.

24. Estevez-Abe, Margarita, Torben Iversen, and David Soskice. 1999. SocialProtection and the Formation of Skills: A Reinterpretation of theWelfare State.

25. European Commission. 1997. Impact on Manufacturing: PharmaceuticalProducts., Vol. 2. The Single Market Review. Luxembourg: EuropeanCommunities.Notes: Label: drugs

26. F&C, Towers-Perrin. 1992. L'industrie phramaceutique. Paris: Syndicatnational de l'industrie pharmaceutique.

27. Faibis, Laurent. 1984. Les laboratoires pharmaceutiques. Paris: DAFSAKompass.

28. Finkelstein, Stan N., and Peter G. Bittinger. 1993. Price Controls and theCompetitiveness of Pharmaceutical Firms: A Preliminary Look at theExperience of Five Countries. Sloan School of Management,Program on the Pharmaceutical Industry, Working Paper 8-93.

29. Geddes, Andrew. 1991. Free Movement of Pharmaceuticals within theCommunity: The Remaining Barriers. European Law Review 16, no.4: 295-306.

30. Gholz, Eugene. 1999. Getting Subsidies Right: Government Support toHigh-Tech Industries. MIT Department of Political Science PhDDissertation.

31. Grabowski, Henry, and John Vernon. 1994. Innovation and StructuralChange in Pharmaceuticals and Biotechnology. Industrial andCorporate Change 3, no. 2: 435-49.

32. Grabowski, Henry G. 1981. Regulation and the International Diffusion ofPharmaceuticals. The International Supply of Medicines:Implications of U.S. Regulatory Reform. Edited by Robert B.HelmsWashington, DC: American Enterprise Institute.

44

33. Grabowski, Henry G., John M. Vernon, and Lacy Glenn Thomas. 1978.Estimating the Effects of Regulation on Innovation: An InternationalComparative Analysis of the Pharmaceutical Industry. Journal of Lawand Economics 21, no. 1: 133-63.

34. Greenwood, Justin, and Karsten Ronit. 1991. Pharmaceutical Regulation inDenmark and theUK: Reformulating Interest Representation to theTransnational Level. European Journal of Political Research 19: 327-59.

35. Hall, Peter A., and David Soskice. 1999. An Introduction to Varieties ofCapitalism. Varieties of Capitalism: The Institutional Foundations ofComparative Advantage. Edited by Peter A. Hall, and DavidSoskiceDraft.

36. Hancher, Leigh. 1992. Competition and the European PharmaceuticalMarket. The Antitrust Bulletin: 387-428.

37. ———. 1989. Regulating Drug Prices: The West German and BritishExperiences. Capitalism, Culture, and Economic Regulation. Editedby Leigh Hancher, and Michael Moran, 79-108. Oxford: ClarendonPress.

38. ———. 1990. Regulating for Competition: Government, Law, and thePharmaceutical Industry in the United Kingdom and France. Oxford:Clarendon Press.

39. Hancké, Bob. 1999. Revisiting the French Model: Coordination andrestructuring in French industry in the 1980s. working paper FS I 99-301, Wissenschaftszentrum Berlin.

40. Hirsch, Paul. 1975. Oganizational Effectiveness in the InstitutionalEnvironment. Administrative Science Quarterly 20: 327-38.

41. Hofstetter, Sandrine. 1996. "Les restructurations dans l'industriepharmaceutique dans les années 1990." Dissertation, PantheonSorbonne.Notes: Academic Department: University of Paris 1

42. Hollingsworth, J. Rogers. 1997. Continuities and Changes in SocialSystems of Production: The Cases of Japan, Germany, and theUnited States. Contemporary Capitalism: The Embeddedness ofInstitutions. Edited by J. Rogers Hollingsworth, and Robert Boyer,265-310. Cambridge: Cambridge University Press.

43. Hollingsworth, Rogers, and Robert Boyer. 1997. Contemporary Capitalism:The Embeddedness of Institutions. Cambridge: Cambridge UniversityPress.

44. Howells, Jeremy, and Ian Neary. 1995. Intervention and TechnologyInnovation: Government and the Pharmaceutical Industry in the UKand Japan. London: Macmillan Press.

45. Jaeger, H., and R. Kaukewitsch. 1998. Le droit pharmaceutique enRépublique fédérale d'Allemagne. Droit Pharmaceutique. Edited byJean-Marie Auby, and Frank CoustouParis: Litec.

46. Juès, Jean-Paul. 1998. L'Industrie Pharmaceutique. Paris: PressesUniversitaires de France.

45

47. Kaitin, Kenneth. 1996. Pharmaceutical Innovation in a ChangingEnvironment. Contested Ground: Public Purpose and PrivateInterest in the Regulation of Prescription Drugs. Edited by P.DavisNew York: Oxford University Press.

48. Kidd, Dan. 1996. The International Conference on Harmonization ofPharmaceutical Regulations, the European Medicines EvaluationAgency, and the FDA: Who's Zooming Who? Indiana Journal ofGlobal Studies 4, no. 1: 183-206.

49. Kitschelt, Herbert, Peter Lange, Gary Marks, and John D. Stephens. 1999.Continuity and Change in Contemporary Capitalism. Cambridge:Cambridge University Press.

50. Koekijk, Kees, Zhen Kun Wang, and L. Alan Winters. 1996. MarketOpening, Regulation, and Growth in Europe. Economic Policy 23:443-67.

51. Kogut, Bruce. 1991. Country Capabilities and the Permeability of Borders.Strategic Management Journal 12: 33-47.

52. Levy, Jonah D. 1999. Tocqueville's Revenge: State, Society, and Economyin Contemporary France. Cambridge, Massachusetts: HarvardUniversity Press.

53. Matraves, Catherine. 1998. Market Structure, R&D and Advertising in thePharmaceutical Industry. WZB Working Paper FS IV 98-17.

54. Milner, Helen. 1988. Resisting Protectionism: Global Industries and thePolitics of Global Trade. Princeton, New Jersey: Princeton UniversityPress.

55. Mintzes, Barbara, and Catherine Hodgkin. 1996. The Consumer Movement:From Single-Issue Campaigns to Long-Term Reform. ContestedGround: Public Purpose and Private Interest in the Regulation ofPrescription Drugs. Edited by P. DavisNew York: Oxford UniversityPress.

56. Möbius, Klaus, Ekkehart Seusing, and Adolf Ahnefeld. 1976. DiePharmazeutische Industrie in der Bundesrepublik Deutschland:Structur unf Wettbewerb. Tübingen: J. C. B. Mohr.

57. Mobley, Jane. 1990. Prescription for Success: The Chain Drug Store.Kansas City: Lowell Press.

58. OECD. 1985. L'Industrie pharmaceutique: Questions liées aux échanges.Paris: OECD.

59. Orzack, Louis H., Kenneth I. Kaitin, and Louis Lasagna. 1992.Pharmaceutical Regulation in the European Community: Barriers toSingle Market Integration. Journal of Health Politics, Policy and Law17, no. 4: 847-68.

60. Plat, Monique. 1970. Droit Pharmaceutique industriel., Vol. 1. Paris: Centrede documentation universitaire.

61. Polanyi, Karl. 1944. The Great Transformation.62. Porter, Michael. 1990. The Competitive Advantage of Nations. New York:

The Free Press.63. Pujol, Rosemonde. 1978. Une antenne pour les consommateurs au

46

ministère de la Santé. Le Figaro.64. Reich, Michael R. 1990. Why the Japanese Don't Export More

Pharmaceuticals: Health Policy as Industrial Policy. CaliforniaManagement Review: 124-50.

65. Sargent, Jane A. 1985. The Politics of the Pharmaceutical Price RegulationScheme. Private Interest Government: Beyond Market and State.Edited by Wolfgang Streeck, and Philippe C. Schmitter, 105-27.London: Sage Publications.

66. Saura Estapa, Eduard. 1994. "L'accord-cadre de l'industrie pharmaceutiquede janvier 1994: une event study." Thesis, Institut d'Etudes Politiquesde Paris.

67. Schatz, Heribert. 1983. Consumer Interests in the Process of PoliticalDecision-Making: A Study of Some Consumer Policy Decisions in theFederal Republic of Germany. Journal of Consumer Policy 6: 381-95.

68. Scherer, F. M. 1993. Pricing, Profits, and Technological Progress in thePharmaceutical Industry. Journal of Economic Perspectives 7, no. 3:97-115.

69. Schwartzman, David. 1976. Innovation in the Pharmaceutical Industry.Baltimore: Johns Hopkins Press.

70. Seillier, M. Bernard. 1991. Rapport fait au nom de la commission desAffaires sociales sur le projet de loi relatif à la pharmacie d'officine.Raport 257 d la séconde session ordinaire, 1990-1991.Paris: Sénat.

71. Shohet, Simon, and Martha Prevezer. 1996. UK Biotechnology: InstitutionalLinkages, Technology Transfer and the Role of Intermediaries. R & DManagement 26, no. 3: 283-98.

72. Shonfield, Andrew. 1969. Modern Capitalism: The Changing Balance ofPublic and Private Power. London: Oxford University Press.

73. Sitzius-Zehender, Hennelore, Bertram Dervenilch, Frank Diener, and GerdM. Foh. 1992. Federal Republic of Germany. InternationalPharmaceutical Services: The Drug Industry and Pharmacy Practicein Twenty-Three Major Countries of the World. Edited by Richard N.Spivey, 227-50. New York: Pharmaceutical Products Press.

74. Smith, Christopher R. 1995. Pharmaceuticals, Intransigent Member Statesand a Single Market. Boston College International and ComparativeLaw Review 18, no. 2: 471-83.

75. Soskice, David. 1999. Divergent Production Regimes: Coordinated andUncoordinated Market Economies in the 1980s and 1990s.Continuity and Change in Contemporary Capitalism. Edited byHerbert Kitschelt, Peter Lange, Gary Marks, and John D. Stephens,101-34. Cambridge: Cambridge University Press.

76. ———. 1996. German Technology Policy, Innovation, and NationalInstitutional Frameworks. Wissenschaftszentrum Berlin, DiscussionPaper FS I 96-319.

77. Storper, Michael, and Robert Salais. 1997. Worlds of Production: The ActionFrameworks of the Economy. Cambridge, Massachusetts: HarvardUniversity Press.

47

78. Streeck, Wolfgang. 1996. Lean Production in the German AutomobileIndustry? A Test for Convergence Theory. National Diversity andGlobal Capitalism. Edited by Suzanne Berger, and Ronald Dore, 138-70. Ithaca: Cornell University Press.

79. ———. 1992. Social Institutions and Economic Performance: Studies ofIndustrial Relations in Advanced Capitalist Economies. London: SagePublications.

80. Syndicat national de l'industrie pharmaceutique. L'industrie pharmaceutique:Ces réalités. Paris: SNIP.

81. Tarabusi, Claudio Casadio. 1993. Globalization in the PharmaceuticalIndustry: Technological Change and Competition in a TriadPerspective. STI Review 13: 123-61.

82. Tarabusi, Claudio Casadio, and Graham Vickery. 1996. La mondalisation del'industrie pharmaceutique. La Mondalisation de l'industrie: vued'ensemble et rapports sectoriels.Paris: OECD.

83. Tortajada, Stéphane. 1996. "Impact des réglementations publiques sur lavaleur des entreprises pharmaceutiques: le cas français." Thesis,Institut d'études politiques de Paris.

84. Viala, Georges, and Daniel Vion. 1994. La loi nº 94-43 du janvier 1994 et lapharmacie. Revue De Droit Sanitaire Et Social 30, no. 2: 239-60.

85. Wertheimer, Albert, W. Michael Dickson, and Becky A. Briesacher. 1996.Pharmacy in the Western World Health Care Systems. ContestedGround: Public Purpose and Private Interest in the Regulation ofPrescription Drugs. Edited by P. Davis, 159-76. New York: OxfordUniversity Press.

86. Wieviorka, Michel. 1977. L'Etat, le Patronat, et les Consommateurs: Etudedes mouvements de consommateurs. Paris: Presses Universitairesde France.

87. Ziegler, J. Nicholas. 1997. Governing Ideas: Strategies for Innovation inFrance and Germany. Ithaca, New York: Cornell University Press.